I'm currently on a marathon to keep my M1 Pro running until the M7 Pro/Max is available. Not sure if I'll make it, but I'm not eager to pay $6000+ for a decent spec dev computer, until I really have to
Have an M1 air as well. Got a Pro m5. For daily coding tasks, and general web browsing the M1 air still holds. I'm going to pass it on to my kid. Reminds me of the honda we have that we'll also pass down.
The M series MBAirs, and ofc the M1 originally, are hands down THE BEST laptops ever made. I have a maxed out MBP ofc but the Airs are unbeatable quality and price/performance.
I think the M1+ MacBook Airs will be regarded eventually like the original Macintosh originally the iPod, or the introduction of the original Air and the creation of the product class that came to be known as “ultrabooks”. It really changed everything about laptops to have a fantastic keyboard, speakers, screen, trackpad, hinge, and 12+ hour battery life… for $1k.
Biggest issue with my M3 Max is apple mandatory fun bloatware. I think it's the same playbook as phones long term. The hardware is still powerful but the software is optimized for the latest CPU, so older models run hot, take a fraction of the second longer to load the resources, etc. That stacks up over 5-6 years of releases to make older models feel slower.
I’ve been a Mac user for 20 years and it’s wild to think how slow my 2003 PowerBook G4 or even my 2008 MacBook Pro felt after five years compared with my 2021 M1 Pro.
My 2013 Core i7 took almost eight years to get to where the previous ones were at five, and if it weren’t for local LLMs pushing me toward an upgrade in the next year or two, I have to imagine my M1 Pro could be my daily driver well into the next decade.
My 16" M1 Pro personal laptop is literally just as smooth for 99% of tasks as my work M5 pro. It's crazy how well they built these. I mean the screen on this and the speakers are superb. And you can find these for like 500-800 bucks 2nd hand. I just hope they give us a few more years of MacOS updates.
I'm wondering what decrease in performance you were expecting from a five year old computer? It isn't like the processor suddenly gets slower simply because it's older.
I rode a 2009 MacPro5,1 (¡galliently! into the future, with capable GPU upgrades) until Jan23's M2Pro mini debut (which I still use as my main Mac).
Apple Silicon is certainly incredible (I also have a base-model M4mini, but it severely lacks RAM/ports so I don't really use it much/eWaste); my only Silicon-purchase-regret was buying a beautiful new 2024 MacBookAir (15") – heavily discounted $850@CostCo – but with only 8GB of RAM (it is functionally worthless, other than for watching it destroy your SSD with page-outs as you casually stream YouTubies...). In 2024, only 8GB of RAM (soldered onto the board because fuck you)!
Macbook Pro 14" Max M1 (64GB). Best Apple device I've owned and worth the investment at the time, especially given the prices now. It's still pretty great for being as old as it is. Have no intention to pay inflated prices, I'll ride it out.
How useful would that actually be for local inference, when the bottleneck is, essentially, always memory size? I think the sweet spot is somewhere around 64 to 128 GB. That ought to be enough for almost all uses, like Qwen3.8-Flash-Next at a decent quant.
Memory size is a bottleneck, but I have found that speed is very important for getting actual work done. Running a big model is great, but if you're getting 10t/s you're going to be spending a lot of time waiting.
I'm loving my M1 Macbook Air still, and only a few weeks away from six years of use Sometimes it lags a bit when I'm running a macOS VM, and all the coding agent VMs, so I'm kind of interested in going to 32GB of RAM, but honestly this computer is still faster than any of my laptops that came before, even with the visual atrocity of Liquid Glass.
I'll go and look at this new box, drool a little, and realize that I'd get more personal utility from donating $3k (or whatever the trade in delta is) to a charity than upgrading my computer.
Been on HN for many years now, and the one thing that never changes is people posting angrily about small UI changes. It makes me wonder- if things have only ever gotten worse, what was the peak UI? MS DOS?
Small improvements don't get much accolades. And for the most part, the UX is so good that most changes tend to be neutral or for the worse. And when you're really attached to something and there's on an unnecessary change, it feels almost like betrayal!
I, for one, prefer Apple products because they sweat the small stuff more than the other options, and when they have put in effort to make things worse it's kind of infuriating.
I do a bit of desktop Linux work too, and it's OK, but hard to complain when it's all open source volunteer work! Windows is so terrible that I wouldn't even know where to start. What's the point? There's nobody at Microsoft that even gives a damn.
Leopard was better because it introduced spaces. Honestly I think that whole concept (I know windows has something analogous now) was the biggest step change in OS functionality I've seen in my life having started on windows 95.
Apple was actually last to the party on that. Linux had virtual desktops for years at that point, and Windows had them as well through Microsoft PowerToys. I’m pretty sure I was running it on Windows 2000.
I had virtual desktops on UNIXes with X11 well before Linus started working on Linux. They came for free to Linux because they were implemented by many window managers.
Not Linux, the X Window System. The first implementation in a window manager predates Linux by a couple of years. And to no surprise Xerox had them first
The problem is that window management is so atrocious in Mac OS that using multiple desktops is just asking for trouble. Like of you close one window of some software, it will jump to some random other window of the same software on some other desktop.
In the end you’ll spend a significant amount of your time trying to “get back” to where you just where (that alt-tab, a possible savior for this in all other OSs, doesn’t work at all on Mac just adds to the UX mess)
Ehh I'm not to caught up with window management on macos. Bunch of shells open I might use tmux. But otherwise I just keep things in piles organized in spaces, and use expose all to get to whatever I need in the pile. It works much like the piles on my desk in real life so I'm comfortable with it. But it was better in leopard when it was a grid of spaces vs a line of spaces. Less rigidity was nice.
It doesn't make a lot of sense to me as a default, but there's an option to turn off the automatic switching of spaces when switching applications, and it pretty much eliminates the unwanted switches from closing the last window on the current space.
Eh. I get what it's going for. The Multi-Window application concept and what not.
In the end though I just feel like it's extra cognitive load to have to think about two different shortcuts to do what is in many ways the same thing.
Say, I've got two browser windows open and a code editor open. I'm wanting to flip through the browser windows to check on things while I work in the code editor. In reality I'm just wanting to juggle the three windows, but depending on if I'm wanting to go from browser to browser it's a different shortcut than if I'm wanting to go from editor to browser. And there's a good chance I'll then need to swap from editor to browser, then issue a different shortcut to go from that browser to the other. When I could have just hit the same shortcut twice.
I've just never really encountered a situation where I've felt like it added something, but I'm constantly having cases where its more annoying.
Thanks, that does make sense. Though as a many-windower alt-tab is just often tooooooo big for my tastes to even try to rely on.
The 40-year-deep-history Mac answer to workflows like that with a small working set with frequent cross-app-one-to-other-switching IMO is mouse-based navigation + spatial partially-overlapping windows, but this seems to be an extremely foreign concept to Windows/*nix users that grew up on full-screen-maximize or tiling. So I don't think is particularly likely to be adopted by anyone who doesn't already have it burned into their brains at this point. :) There's a lot of these old Mac bones still baked into the OS (e.g. if you have a folder on your desktop, and double click it, and move it, then close it, it'll open to the location you moved it to next time. And that memory is separate from that for another folder on your desktop. Though some of the follow-on behaviors have changed (opening folders in already-open windows used to behave similarly, opening new windows by default).
The 20-year-history version is my own personal lord and savior: Exposé hot corners.
Spaces and no-mouse navigation are still in many ways comparative afterthoughts to mouse-based/spatial workflows. Though as someone with 30+ year of Mac muscle memories in my head, the switching-windows-vs-switching-app concept and awareness of "where" I am relative to it is pretty embedded.
Firstly backtick doesn't go to the most recently used.
The whole window arranging so you can click on them doesn't help you when you have multiple desktops, close some terminal, and jump to some other terminal three desktops away. Like, arranging windows so you see some part of it always, just so you can click on it to switch back and forth between windows feels like a horrible workaround for broken alt-tab behavior. But there's additional levels of weirdnesses that just make using Apple frustrating and scary, like it's constantly throwing random curve balls. It just reeks disorganization. Which is really unfortunate - 10 years ago or so (when witcher task switcher still worked) I really enjoyed working mac. Back then I was mostly coding in the same windows I guess. Nowadays it's business work, and it's just a pain.
For me it is that cmd+` only cycles between windows of the same app in the active desktop. It makes it hard to switch back to a window of the app that was put in full screen.
I get that that behaviour shouldn’t just be changed but I would appreciate it if there were another key combination that would allow cycling through all windows of an app.
Liquid Glass was not a "small UI change". It disrupted many accessibility aspects of macOS, so much so that Apple toned down quite a few aspects of the design in subsequent updates, even adding a slider to reduce the effect. [1]
The quantity of requests for how to disable Liquid Glass on Apple Discussions is staggering. [2]
Yea, I am trichromatic color blind, and it is legitimately unusable for me. I am unable to type on the keyboard. I know there are changes since it launched, by after my ipad updated automatically, rendering it near useless to me, I just disabled updates on my iphone.
I think the biggest thing isn't really whatever it is but the act of changing and forcing everyone to change into the new ui that is the disruptive part. It is a combination of window dressing changing but also sometimes functions breaking. It is ironic because software written for developers prides itself generally on maintaining compatibility with say depreciated flags, or if things will be depreciated plenty of notice ahead of time. But that only holds for cli software. GUI software and the OS that all goes out the window. Stuff breaks, workflows break, you have no recourse and in fact they might make it genuinely difficult to go back to an older version. Like this part of it is the shitty part. Never mind how it happens to look. It is the forcing effect.
I think it’s also because these UI changes are always designed to improve first-impression and first-usage. They are outright hostile to anyone who wants to go fast and get faster — the power user. At minimum the re-arrangement resets any gains you had, and at maximum its irrecoverable because the UI is now fundamentally designed to slow you down.
This eventually becomes awkwardly precise complaints about information density and general change; but the underlying issue is I want to go fast and the system says no
I think the only real solution to this is to switch to an OS with a backing org that has optimization for power users as a first-class goal. I've been liking Omarchy and NixOS for that. I say this as a Mac user since 1989.
It's been a long time, but I'm pretty sure Appearance Manager and the corresponding unreleased-but-leaked Hi-Tech theme was 8.5-or-so-and-up-only, which means I must reject 7.6 on aesthetic grounds ;)
On the Windows side, Windows 2000 and XP/Vista/7’s Classic mode, though Aero was quite nice. I’m not a fan of the interfaces from 8 onward, and I didn’t like Luna in XP.
On the Mac side, it’s a toss up between the platinum interface of Mac OS 8/9 and Mac OS X Tiger. With that said, I have a work-issued MacBook Pro running Golden Gate as my daily driver, and Liquid Glass is alright. It’s not what I’d prefer if I had a choice, but it’s not a dramatic departure from previous versions of macOS. A Cheetah or Puma Mac OS X user should be able to navigate Golden Gate with ease, mostly.
I am a HUGE sucker for Mac os8/9. I currently have a PPC Mac Mini and a Powerbook G4 that I tool around with occasionally for nostalgia's sake. I used slew of various G3 computers running OS9 as my main personal computers picked from the scrap bin at my dad's job (graphic imaging company) from ~2004 until I went off to college in 2010. I had some kind of modified Mozilla browser (Classilla I think?) that had modern-ish web support.
I've been on the lookout for a good condition G3 beige desktop, that was my first ever computer that was "mine" and I really want one just for the sentimental value.
If you consider the transition from macOS 15 to 26+ 'small', I think you should see an optician.
Like GP, I despise the new design. I'll say the old design was pretty peak design. Windows looked like windows. Buttons like buttons. With accessibility tweaks (for transparency and borders) I could make it very, very usable.
This new thing is a crime against UI. Every icon now sits in a circle, and takes up % more room as a result. Because … it's pretty? It isn't pretty.
I upgraded my laptop thinking I was going to develop an app and that I'd need to be on the latest design. I ended up developing a TUI so I needn't have bothered. Every time I see my partner's laptop, still on Sequoia, I shed a little tear.
Perhaps it is the annoyance of change for change's sake. If you buy a car and then the interior changes after an update, it'd be pretty annoying, yet that is what is foisted on us.
Liquid Glass is more than just the transparency. There are many other changes that cannot be disabled that way. Not to mention the overall lag compared to Sequoia (I skipped Tahoe entirely so I had a direct knife-edge transition). 27 is less bad than 26 but Liquid Glass doesn't make anything better than prior UI and makes several things worse, even if you turn down the transparency.
I imagine that the UI changes since Sequoia regarding the giant buttons and whitespace are because Apple plans to release a touchscreen laptop (most rumors say it will be announced tomorrow) so they've decided to make the UI touch friendly at the expense of bloating the UI for everyone who doesn't have a touchscreen. Microsoft made this same decision with the Windows releases after 7.
Why? My SO has a Windows laptop with touchscreen, touchpad, keyboard and mouse. She uses the mouse but sometimes pushing a button by touching the screen saves time. If you think about it, by the time you reached for the mouse and started aiming for the button you already pressed it with your finger. The touchpad reduces the travel of the hand but not the time to aim. Hitting something with a finger is always faster. She is missing that on her work laptop.
> If you think about it, by the time you reached for the mouse and started aiming for the button you already pressed it with your finger. The touchpad reduces the travel of the hand but not the time to aim. Hitting something with a finger is always faster.
Depends on how accurate you need to be. How fast/confident you can be at hitting with a finger can easily drop off a cliff.
> The touchpad reduces the travel of the hand but not the time to aim.
This is likely true for casual users, but for veterans this is a skill issue. My touchpad accuracy/speed more than matches most folks touchscreen accuracy/speed.
It may also reflect the quality of the touchpad. For a Windows laptop, it's virtually guaranteed to be of lower quality than something offered on an Apple laptop.
Reaching up to touch the screen is slower than reaching down to the trackpad. Also consider trackpoint.
Skilled users can make the reach and the motion one. You don't need to stabilize your palm on the rest to use the trackpad. Same way I don't need to use F and J for the keyboard (and kinda wish I could remove those annoying bumps)
I give you an example on this page of comments, so you can check if my informal results apply to you too.
I am moving my finger on my touchpad to move the pointer to more or less match what I am reading. I am always close to a Reply button or a vote one. There is no point reaching for the screen with my hand and touching those links. Furthermore they are way too small to be pressed with a finger. They are probably too small for a pointer too, but that's another issue.
Hoever, when I open an empty tab, type new in the URL bar and FF completes it to the URL of this site, and I press return, I usually end up with a Login link in the top right. If I reload the page (ctrl-r) I am logged in. For some reason either FF is not sending the session cookie or HN is not doing what it should on a cold session. This is also another issue (sorry) but let's say that I want to login. My pointer is far away from the login link. My fingers are not on the touchpad. Is it faster to move my right hand towards myself, touch the touchpad, slide the pointer to the link or moving that very same index finger to the link and touch it? My laptop does not have a touchscreen but when I used a tablet with a mouse and a keyboard I reached for the screen many times. Of course a mouse is less efficient than a touchpad, except for games, but I also miss the touchscreen sometimes.
And I would use if to press the reply button below this line I am typing. Let's see where the pointer ended up while I was typing... At the top right corner. Moving it to Reply and clicking is slower than touching the screen.
I used touch on Windows to test multitouch and pen input support in a familiar environment. Less hassle than grabbing a tablet, setting up remote DevTools, etc.
The way things are going you will be able to run your favorite flavor of Linux on those pretty soon.
It will be the best of both worlds: the best hardware and a good OS. Hopefully Apple doesn't take steps to block it.
I got a second hand M1 air like two years ago for 200usd with 1tb hdd and 16gb ram. I got it after my intel Air died after about..12 years? (Edit: the ssd died, which is removable… but literally had a plane to catch). All my pros have died within a few years. Air all the way. Never looking back.
I wonder if this is from not trying to get maximum performance, and instead focusing on efficiency? But I've noticed this too. I definitely prefer the Air form factor over larger models as the primary benefit, but the durability seems to be a big benefit too. Fewer ports, thinner shell that seals tighter and can't flex as much due to the curved metal instead of flat surfaces... everything about it just seems like a better design.
No fans though. I like being able to crank them full blast. I'm not sure this pro would ever actually run the fans all out with my usage with the existing fan curve. Apple always favors it running quiet and hot, but I don't mind loud and prefer cool palms. I grew up with far more noisy rigs.
Just use. There were two models that had issues with the gpu desoldering, 2012 and 2015 iirc? I was lucky enough to have them. I put the mobos in the oven (180deg for seven minutes?) and that would fix it. For about a month… and then 2 weeks.. and then… I actually have a 2018 that I borrowed from a friend after my 2013 Air ssd gave up, it is still working I think, but its twice as slow, at least as my Air was. Its so weird. The Air stayed ‘snappy’ even after my friends Pro’s would become too sluggish. Maybe because I never update ;)
Did they really wait that long? I thought most commercial products went lead-free in like the '80s or '90s. I have an old Sony PVM with known issues due to low quality lead-free solder, fixed by reflowing/adding fresh solder. From what I can gather it was released in 1994. In the same vein I'd have expected lead-free to be a bit better by 2012.
At least two of these are now on completely unsupported OSes that are no longer receiving security updates (and lots of latest versions of apps won't work). Especially given the uptick in major CVEs of late I'm very nervous about reconnecting any of these machines to the internet :/
no one is going to hack your laptop that sits behind your router at home checking emails and browsing the web
I get that this is hn and we're paranoid about everything, but come on, what CVE is a threat to joe sixpack here where he should drop $2K on a new machine to protect against it?
That's what Linux is for I've got a 2011 MacBook Pro running Debian and its still perfectly usable(after maxing it out to 16 GB RAM, installing an SSD, tossing the CD drive, repasting the CPU, and replacing the battery). I'm curious to see how long it can keep on trucking, at some point the charger may be the difficult part to procure; it’s not very robust.
It's a great point! The 2015 has been a linux machine for ~5 years, and the 2012 is my wife's laptop, used once every couple of weeks. Trying to get her to upgrade to a MacBook Neo for this exact reason.
In 2024 I sold a 2010 MBP that still worked. I beat the hell out of it in college dual booting and cooking the cpu playing Trackmania for 12 hours at a time or running Solidworks, then after I graduated it sat closed behind my stove sharing its wifi connection (my neighbors unprotected wifi) out the ethernet port and into my own managed router for about 4 years. Up until the last days it was running Windows as a host for controlling my old not-network-connected 3d printer. I definitely got my money out of that machine, someone paid $150 for it.
The last few models of Intel Mac Pro that I used were mostly good at getting really hot.
Dropbox would fire up to do massive walks of the filesystem tree and the fans would take off. I think the thermal interface material on the GPU got cooked at some point. It would be a relatively easy fix, for me a couple of hours, but I chose to fix it with an M1 MacBook Air.
The old airs were legendary. My 2012 still runs just fine despite being beaten to hell. Would still be a great dev machine for non-GPU workloads if it weren't for a lack of OS updates. Have been very happy with my M1 Pro as well though.
Wow turns out there's a whole community of us! Same boat. I usually recycle, hand me down, trade-in but have really had a hard time letting that little one go.
You gotta be kidding me. The old M1 Air had the "bendgate" where several people, me included, saw their screen die overnight for no reason. Tens and tens of pages of people complaining about it only to be with "you're holding it wrong" answers. Insanity.
I think Apple fixed that mess with the gen after that one.
I'd still buy MacBook Airs even if they cost more than the Pros. The lighter weight and the silence of having no fan are worth more to me than the extra performance to me.
I think differences were greater pre-M1, these days Airs are like low budget config Pros missing the fan. I highly doubt they'd last any longer. Maybe you could argue that with the best CPU/GPU options used in the MacBook Pro it will degrade faster than any Air, but if you spend about the same on each, I expect the Pro to win.
My son's M1 Air CPU died suddenly. Maybe it was a thermal death from overheating. Not sure. But all the data on it was lost because the storage is soldered on. PSA to make sure devices with non-removable storage are backed up!
(We replaced it with a Framework... with removable storage.)
In Windows, by default, Microsoft tools put projects in ~/source/repos, which OneDrive, in turn, by default won't touch so they'll be lost if the SSD fail and they're not in git.
However, OneDrive doesn't seem to even know how to handle source files (it doesn't even attempt to honor the gitignore), so rebuilding a project in folders covered by OneDrive results in flurries of temporary files being uploaded/updated to the cloud, hogging the bandwidth, with a bonus of locking files, crashing builds, or even reverting changes silently.
One would think backup tools, even if primarily for normal users are still literally made by programmers, at least consider the use case of programmers instead of just ignoring them entirely.
>file sync is backup against the things most people have to deal with
I would say accidentally deleting a file is more common than a house fire, so NO, file-sync is not a backup nor is RAID (block-sync) or clustering (data/state-sync).
A backup is/should be a non-modifiable snapshot of data/files maybe even a state.
I give you an example:
You have a password-safe (as a file). Your computer blows up, but you are safe, because your password-safe is synced on Dropbox....but oh my god what was the password to Dropbox??? Your restore is gone because you forgot to "backup" your password-safe.
However you are not a Loser, being a Admin is a ~lost art and a really good paid one (again).... ;)
It goes to your trash can/recycle bin for a while. That's literally what it's for: to guard against accidental deletes. Very few people accidentally delete files from their computer's drive anymore. The only time people do this in 2026 is accidentally formatting an SD card that had their camera's pictures on it or something.
I guess I'm not tracking. I put all my source folders in my Documents folder on the Mac, and check the checkbox to have that sync to iCloud. It works quite well.
I also do a bunch of video work with Da Vinci Resolve, and those project files are in the same folder. I don't keep the clips there, those are on an external disc, but that's just because I haven't gone from the two terabyte iCloud plan to the six terabyte plan.
Surprising somehow, to me.
One would imagine Apple could make a good amount of money by selling storage and thoughts-off safety for normies?
And catch up to what Google offered like then years ago already with ChromeOS.
Yeah I've always thought it was weird, especially after they cancelled the Time Capsule, leaving no official Apple product to back up your Mac. I figured it was because the max iCloud plan was 2 TB and Apple sold Macs bigger than that, but then they increased the iCloud plans and still didn't do it...
> PSA to make sure devices with non-removable storage are backed up!
Really this should be devices full stop. Also, "backed up" should include no fewer than two backups on different media and in different physical locations e.g. Time Machine to an external SSD _plus_ continuous cloud backup.
What would have been the right amount of harping about the environment though? Or is it always going to be a pointless exercise trying to convince/alert people?
Or is the only way to do that successfully, doing it subconsciously/covertly like the state actors are swaying public opinion? :-/
1. And yet we have someone on HN reminding everyone to even have a backup in the first place so let's not assume everyone on HN can't also make "average people mistakes".
2. You mean the climate activists who used to risk their lives back in the 80s on open oceans instead of gluing themselves to streets? Yeah, the times have changed.
I agree, if it has too much friction it's more likely you will postpone it. Any option, either just a cloud backup tool or a spare HDD or SSD with CCC or TimeMachine is better than nothing. And you have to plan for the most likely case, e.g. spilling beer on the laptop vs the house burning down.
For the house burning scenario, a cheap strategy is having one or two old HDD's in USB cases, located somewhere away from your house, in a place you regularly visit and trust security-wise. Then you can just make a habit of dumping the files there whenever visiting. Most important data of mine is safe even if internet goes down and somebody nukes the city I live in, for zero recurring cost.
While I was expressing an ideal setup, it isn’t all or nothing. A single backup is still better than none at all.
I suppose I didn’t specify exactly what I would recommend to someone who thinks it’s too hard, so here is that:
1. Install Backblaze. This is basically fire and forget.
2. (1) is enough for most people, but if interested (and on a Mac, which is what I’m familiar with), plug in an external SSD once a week or so for Time Machine. Setup involves clicking on the drive in the setup wizard and specifying an encryption password.
I don’t know about “ultra” robust, but this gives you protection from accidental deletions, natural disasters happening to your computer, hardware failures, and corruption from software bugs. The ongoing maintenance is plugging in a hard drive once a week.
If that is too much effort, just (1) provides natural disaster and hardware failure protection. It helps somewhat with the others but Backblaze only saves versions for 60 days and that’s a tight window to notice subtle corruption of e.g. your photos.
if you want to talk about 'average people' go hang out on reddit. How many times have 'average people' been brought up here explicitly to shut down a discussion. It is reductive and insulting
how many good things have 'average people' (and the avaricious wanterpreneurs chasing their dollars) ruined? Nearly everything!
That's a big ask. Probably only work the few hundred it'd take to do that, then get the data off it, if you had some super important documents // crypto wallet.
Is it really? In China you can upgrade your iPhone storage this way. Sure it's not for your average phone case store, but it's not exactly rocket science either.
Unfortunately not really a lot of places to do this in the US. And unlike shipping new devices from China to US, shipping used devices from US to China and back is prohibitively expensive.
I am not familiar with Macs so I might have been wrong from the very start, but if the owner put the key in the secure enclave then the same user should have no problem at putting the same key in the secure enclave of another Mac. If the user can't write into the secure enclave, then the Mac is protecting the data against the very same user that is supposed to own those data. It is protecting against wrench attacks too but what's the most frequent scenario?
The backup would live someplace other than on the broken machine. It would be encrypted some other way (or maybe not at all, depending on threat profile).
No point in encrypting those family photos burned to disc.
Sure, but in this case your cpu dies or ram dies or ssd dies, in all cases you loose data, in an upgradable system you loose data when ssd dies. Different risk factors.
that's just a false sense of security. A machine should be treated as a whole. Separate components only matter economically, not for data security.
My macbook updates my documents etc to my NAS, and the NAS syncs to crucial folders to the cloud (not everything for cost reasons)
You should always assume your device can be stolen, you can spill fluids on it (or a guest), a cat can pee on it, an electric shock may kill it... No matter how great care you take of it, shit happens, data needs to be safe.
The SSD dying standalone vs the system dying doesn't change at all for data security. It's not a different risk factor
Hard drives failing is a far more common problem in my experience. As is theft and loss - one should always treat a laptop as if it will fail at any moment and keep good backups.
If your looking for arguments against soldered storage the facts ssds are wear items and will eventually need replacement is a much better one.
If almost ANY failure means the laptop is basically bricked then even with good backups it means days or weeks before being back up to speed, having a working laptop with recovered data, esp. when working on highly confidential data as agreements can mean you cannot upload the data to a cloud.
If given a choice, I prefer a system where a HDD/RAM/GPU/CPU/PSU/Screen/Port/Battery/Keyboard failure is barely if at all disruptive to one where all of these effectively brick the system.
SSDs are a more common failure point, as is theft or loss or damage.
The point is it shouldn’t matter if the cpu or ssd dies because you have backed up everything already and are prepared for anything to fail or happen to the device rather then “oh the ssd is removable so in these select few situations I can recover my data”
It's unrelated to the physical connection, but the drive encryption used on the built-in drive keeps its keys in the CPU's secure enclave, so dead CPU == data lost, regardless of whether the ssd itself is damaged.
Obviously if you used the same encryption scheme on a removable drive you'd hit the same issue - but most people don't do that, because storing the encryption keys in the Secure Enclave makes your drive effectively no longer removable.
Even if the storage is soldered on the data can still be saved if it's not the storage that died and even then, it could be possible to retireve at least part of the data. Get a data recovery specialist.
I bought my M1 Pro the week it came out and it was unlike any computer I ever had, expectations, compute, battery life especially. It was just a joy to use and not worry about chargers. I'm so glad they went this path.
I had an M1 Max machine with 32GB of memory and upgraded it to an M4 Max machine with 128GB of memory for local AI model experimentation.
It's sad that M5 (and M6) are so much faster at prefilling and other LLM-related operations, but I'm happy with the computer I got and will keep it at least 3-4 generations.
They really have radically improved the situation for prefill speed! But the generation speed is behind enough that you'd notice. An M4 Max is a great choice I think.
Stay strong! I went from an M1 to M5 generation, and it was a nice bump, but I could have gone another year without any significant impact on my quality of life/work.
I have a M1 pro for personal use and an M4 pro from work and I don't notice any difference switching between them. The M1 is still more than enough for most computer usage.
I just bought a m1 Air with 8GB as secondary device to the m4 mbp that my employer provides since they dialed up the spyware lately.
I should have bought a m4 Air last November but I hope the m1 holdes until the rampocalypse is over.
I have an M1 Pro as well but my next computer will be a split setup mainly because I realized that I'm most often working / focusing on hard things while at the desk and just browsing / checking in on things when I'm away from the desk so I'm thinking something like:
that's the plan for now, I just haven't had the M1 Pro slow down/hang/lag enough to give it up. It's a mid-range M1 Pro with 32GB of ram, works for me for dev work, I don't run any local LLM or resource intensive things on it. Battery only has 250 cycles on it so it's not like it dies off-charger.
I have M1 Maxes (one for work, one for home) that perform as good as new, as far as I can tell. Only thing they lack is an Nvidia GPU. I might pick up the new uber-studio though tmw when it launches (presumably).
I've been using my M1 Max 64 GB for 5 years now and it's pretty zippy once you rewrite some slow software. I have an M4 Max with 128 GB but I never ended up migrating to it. That's how good it is.
Not op here, but I rewrote the file manager giving claude only two goals: cold startup below 100ms and 30mb nikon raw preview also below 100ms. One night in a loop and it hit around 20, 30ms respectively. My macs now feel uber fast.
you should, I was surprised how seamless migrating was. I just connected both devices with a usb c cable, waited an hour, and everything was moved. Including application settings etc... It was completely painless. Just needed to login to some new websites etc.., but it took me less than 30 minutes of my effort to migrate, mindblow (total time was 2 hours because it copies all your data, but you don't have to stay with it)
I think I'm on the same computer, the M1 MBP with the touchbar right? I was eyeing an upgrade for a while but now that I've transitioned to running agents on a home server instead of locally, I think I'll be putting my money into a ram upgrade for that box instead. My macbook is basically a dumb terminal at this point, the only reason for an upgrade would be battery life. I might get an air or even a neo if this thing dies.
My M1 MacBook Pro has been the best hardware investment I’ve ever made. I had planned to update it in 2025 at the four year mark, but it’s still going strong so I don’t see a reason. The new plan is to go to the Apple store the very day that it dies.
I also have a high spec M1 and was delaying upgrading, but eventually I needed more RAM.
My M1 is still fast and enjoyable, but I regret to inform you that the M5 actually does feel faster in day to day use. I thought for sure I wouldn’t notice, but the difference is there.
When I do large builds the difference is also bigger than I was expecting.
IMO Apple knows a ton of people are doing this, especially with the rumors on memory bandwidth aka LPDDR6X bringing huge generational uplifts[0], which is why it seems they haven't brought down trade-in values from a few weeks ago, at least for my base spec M2 Pro MBP - Still $835. Pretty compelling to do that and pay just over $100/mo for a year to get a M6, given I've already gotten >3 years out of it.
this machine just keeps getting better and better in the inference world. the memory bandwidth is too good, newer macbook pros aren't that much faster in memory bandwidth, many are slower. not enough to upgrade yet as they're mostly horizontal upgrades, the improvements in other areas are adding up like wifi bandwidth and protocol, NVMe speeds, webcam
if they do the M7 correctly: which is just everything their current models do plus way faster memory bandwidth and much greater memory possible 192GB+, then it will be a really great and timely upgrade
Here are the numbers that got me buying a used M1 Max for LLM work.
M1 Max 400GB/s
M4 Pro 273GB/s
M5 Pro 307GB/s
Only Max are worthwhile for LLMs. An Ultra is just two Mac chips on a die, but I think they're just used in Mac Studio.
I'll check it that inference engine! Thanks!
Edit: it doesn't give numbers for unsloth/Qwen-3.6-35B-A3B 4bit, but I'm using mlx-serve right now and getting petty good numbers (and it took a while to tweak everything properly!)
no it’s stuck there, M1 has a 720p cam with a smaller sensor thats noticeably bad in low light, and the refresh has occurred twice to 1080p with larger sensor and 12MP but cropped to 1080p for audience tracking
That’s what even low spec computers used to cost, not even adjusted for inflation. I don’t think spending $200-400/month on a hardware budget that enables me to earn 50-100x that is that onerous, we just think “six grand for a laptop???!?” But it’s ridiculously cheap for what it is. We wouldn’t be similarly shocked at a $6k car, for example.
I figured my direct family could use an extra Macbook, so I decided to "peel" my M1 Pro away as a gift to my mom. It stays close, I can use her unused space as backup, provide ad-hoc Tailscale to her (way smoother than with her crufty Windows machine), and I get a nice upgrade to my current workflows.
Having said that, I did plan on sticking it out 7 years. Ended up bailing on an M5 Pro!
Also M1 Pro here. Any reason why you're planning to upgrade at M7 Pro? I still don't feel the need to upgrade, as much as I wish I could justify it.
I like to factory reset my devices every couple of years, it really gives them a new lease on life. My M1 Pro feels as smooth and fast as the day I got it.
For the ability to run our full server-side application stack locally, including all services and as much of the test suite as possible, as well as the profilers, debuggers and telemetry tools to debug issues. Way faster to develop when you can have a change running live in about 3 seconds instead of 5 minutes, run targeted tests in seconds and a full test suite in a few minutes.
Not op here, but: i have a build box i ssh to, but when on the go: planes, taxis, etc. its annoying to constantly have to connect. Slightly. Having a fallback, fast local laptop is a convenience. Tho I do travel a lot - usually >100 flights per year.
When I was in the market for a new machine, I had the same attitude. As I was about to hit "Buy" on a standard issue $2k MBP, my wife said, "What if we got you a top of the line machine?"
I replied that it would make no sense unless I was going to start running local inference or training my own models locally...
2 years later, and this MacBook Pro M4 Max w/48GB is the best $4k I've ever spent!
Big test suites. I have one project hitting 35 mins on m1 max, vs around 12 min on the m5 max, vs. 7 min on the linux box I keep in the basement. Unit tests, integration tests, headless browser with webgpu stuff.
Are you on the model with the CD drive or the high resolution "thin body" version? I graded the ram in mine and put in an SSD, and it still works beautifully, but I'm now on now on Linux with it, and have been for some time.
My M1 screen just cracked :(
Debating to get it fixed (it’s a 32GB 2 TB machine). Crazy that I likely will spend 500€ to get a super old machine repaired just so au do not have to spend >3000€ for a similarly equipped new one.
Happily running my M1 Max, but have recently been annoyed at my decision to go with the off-the-shelf 32GB version at the time. I recalled telling friends at the time that the ridiculous memory bandwidth should offset that, so I ran some tests on memory pressure while running all my mess and, well, whaddya know, everything is actually tickety-boo. No memory pressure for the most part. And this is with big rust projects, Orbstack builds and runs.
The engineering that went into the M1 was amazing, I truly think that now, an M1 Max was the biggest bargain ever in my entire computing life. £3,299.00. In contrast, the 64GB i9 I bought only two years before was £4,149.00.
The “on-device AI” with the standard M6 seems to be slightly faster than the M1 Max of five years ago. All the rest seems to more or less match. Hope the M7 Max is yet another leap that serves many users for years to come.
The M1 Max MacBook with 64 gigabytes is still a great machine and fairly affordable on the second-hand market.
(Returning an i9 to wait a few more years on the M1, any M1, was absolutely worth it)
Haha same here. The M1 Pro is such a powerful machine and it still works great.
I did buy a MacBook Air M5 for non work, and I think it’s about as fast as the M1 Pro, but the latter has 64GB of RAM so I can actually work. Will probably request a new one next year though, it’s time.
I’m hopeful mine survives until RAM and SSD prices return to Earth. Unless I drop it I suspect it will. It’s still going strong. Maybe the M8. Hopefully chips and local LLMs have developed to the degree that local inference will be good and fast.
I saw a decent listing for a 16" 16gb m1 pro macbook the other day, and i was seriously considering spending those money.
I have a 16" m4 max from work (36gb) and is an incredible machine (build quality, incredible display, overall feel). I also have a private macbook neo which should be similar to M1 in terms of performance but it's limited to 8gb of memory.
The neo/M1 performances are sufficient for my needs (for private computing) and chances are the M1 pro actually performs better than my neo (if anything for the more abundant memory).
I might to back to the listing and try to haggle the price even lower, considering a new generation of macbook pros has just been announced.
I gave away my base spec (8gb ram, 256gb ssd) M1 air just a few months ago and it's still working fine, I just needed something higher than base spec for work but the new owner of the M1 loves it, for general purpose use it may as well be an M5.
I do wish they had removable storage though, I originally got that Macbook as it was the cheapest way I could try out MacOS and M1, once I decided I liked it, I was stuck with 256gb or spending all that money (and then some) again. This time, I went with 1tb.
my intel to M1 Pro upgrade was like stepping into the future, still on it and my only motivation fro swapping computers is i dropped it from my lap when i fell asleep after a long day with the kids, and the n key hasn't worked for a year
authorized repair shop says 4 days to fix it and it's my only laptop, so i go without :(
Indeed, pricing has gone a bit out of hand. I got the top-of-the-line MBP when I was starting my PhD at the end of 2013, for something like $3,200. You couldn't get a more expensive consumer-oriented laptop at the time. It's mind-boggling that the equivalent purchase today costs around twice as much.
Me too, running a 16 inch M1 Pro. So glad I went for 32GB as that'll add to the longevity. Providing no hardware failures I can see at least another 4-5 years. It's been running silky smooth up to now even with the Golden Gate update.
I mean, you don't have to. Do you need a Mac? I make mobile apps and Xcode is only on macOS so I definitely need it but most people don't, there are Windows and Linux options that are great for fractions of the price.
I'm always surprised that devs need such high spec machines just for developing software. Unless you're developing games or scientific computing software or similar, I din't know what the power would be used for.
My dev machine is an i5 of some description about 6 years old, that's about all I could tell you off the top of my head because I've never worried about needing slightly faster compiles. (Ignoring running local LLMs of course. Currently they can't compete with Opus 5.5)
I'm developing an inference engine and need access to the latest Neural Accelerator features, and therefore need to keep up to date with the Apple releases. M1 through M5 Max 128GB. As a broke bootstrapped un-funded startup founder this is extremely tough to afford given the price increases and all.
I'm still running a 4 year old M1 Max 32GB Mac Studio as my daily and really haven't found anything that it shows sign of slowness/age with.
That said I don't do any local AI or 8K video editing type stuff.
On upgrade I could probably live on a moderate/high spec Mac mini since the only reason I went Studio originally was the resolution/number of displays it could support.
I'm with you. Typing this on a 2021 m1 64gb. I keep thinking it's time to get a new one and I'd really like 128gb but justifying the spend is REALLY hard. I've had m5/128gb for a few weeks at a time and it's an upgrade for sure, but not $6k+ upgrade.
I switched entirely over to Asahi Linux on my M1 Macbook Air and installed a third-party 55Wh battery, as the original battery had fallen below 80% capacity. That machine is good to go for another 5 years.
There are no features or performance characteristics of any of the subsequent M[2-6] Macbook Air or Pro models that I care about. I'm not really sold on running models on the hardware I have right in front of me. So long as I can rent an RTX PRO 4000 Blackwell at about half a buck per hour through a European provider, I'm fine using that with Qwen3.8-27B UD-Q4_K_M for the few things I consider more privacy-sensitive.
How is Asahi Linux from a reliability perspective? Last I checked there were some issues with docking and HDMI out that would've made it tough for me to use as a daily driver.
Not a Mac user. I assume HDMI might be improving soon as AMD finally got the HDMI consortium to wind their necks in and full HDMI support was recently merged.
I've been running Ubuntu on an MSI Ge66 gaming laptop, Intel + Nvidia gpu for a couple of years now. I did have some teething problems with wifi and sleep but after resolving those it's been rock solid. I'd say 99.9% stable if we're throwing out random percentages here.
While not Asahi/M1, I can tell you my desktop has been Linux Mint for about 5 years, including through distro upgrades, etc and I have had to troubleshoot it like once.
97% is a high estimate based on my linux efforts.. just yesterday some kind of update, i think kernel, took out my amd graphics.. still wouldn’t pick anything else.
Funny you say this. I had my issues with Linux during the years but last 2-3 years I have been running it with no issues whatsoever on a Dell XPS 9xxx series. I recently got an M5 Max and it had more issues than the Linux/Dell duo.
It wasn't meant to be bait. I'd really like it if Linux was as rock solid as Mac or Windows. But it clearly isn't and it doesn't help anyone that so many Linux fans pretend it is (look at the replies!).
That said, I do have hopes that AI will basically solve this. It's extremely good at diagnosing low level bugs, reverse engineering drivers and so on. A silver lining of AI taking our jobs at least!
> I'd really like it if Linux was as rock solid as Mac or Windows. But it clearly isn't and it doesn't help anyone that so many Linux fans pretend it is (look at the replies!).
It is perhaps also possible that others have different experiences with both Windows/Mac and Linux than you do. It's common for people to not notice what's broken, or what they have to workaround, with the software they use every day. People also have different workflows and priorities that determine whether their use of a tool is "rock solid" or not. (Something I certainly wouldn't say about Windows—or Mac!)
Funny that Linux is perceived as unstable, I have to fix a Windows laptop where Windows starts up fine 1 time out of 10, while normally it freezes (not even blue screens) on the boot loading spinner, I've done everything possible (including of course running a 2 hours RAM test, and testing it with a Linux live USB) to exclude hardware problems, so what is remains is... Windows of course.
One of my coworker that uses a Mac has an issue with a USB dongle that stopped working randomly, not even with a system update. If he uses on its laptop the external monitor remains blank, the same dongle with another monitor and another identical Macbook with the same macOS version works, the same dongle with the same PC and another brand of monitor works, the same PC with the same monitor and a dongle of a different brand does work, but a different dongle of the same model doesn't. So I guess for some reason macOS saved the ID of that dongle and the setting got corrupted. Tried to reset the hell of it and it doesn't work.
I've never had similar issues with Linux. Most of the time when something doesn't work with Linux you just look at the logs, figure out what is wrong, correct it, and it start work fine again. Configuration of the system is predictable, if you install a software and then you uninstall it the system is left in the same state as before installing it, it doesn't leave files that waste disk space, as well as when you upgrade the system (yes, every Windows or macOS update leaves old stuff behind that it's impossible to delete, effectively "leaking" disk space at the point you have to perform a clean install).
Also really happy with Omarchy on Macbook.
I installed the nightly kernel that has usb-c and it works really well. Connects fine to both LG Ultrafine 5k, and Studio Display. https://x.com/AGTPinsights/status/2106564829532328331
Still some tiny issues with waking from sleep lid closed, but I'm expecting to get that fixed myself with some prompting this week. The people working on Linux on Mac are really impressive (and have access to unlimited tokens), so I expect all issues to be fixed very soon.
Being able combine agents, linux software and Apple hardware is so much fun.
It's quite stable. Been running NixOS on my M2 Max with basically no issues (HDMI working fine as well). However the lack of adequate sleep / no support for suspend to disk led to me buying a Framework 13 Pro, and I'm now selling the M2.
I think the point the OP was making was that you can rent this for a couple of hours when you need it, and then it makes sense. If you are using it say 2 hours a week (or even 5 hours a week) it's going to be much cheaper than building your own hardware for a long long time.
It's a little confusing from the page, but you can't actually use those hourly in terms of spinning them up on-demand for a few hours, because you pay the € 209 - 1499 setup fee when you provision one. They're just showing the monthly price broken down into an hourly price.
They're great servers, though, I have a whole fleet of them.
How is using an external provider any better for privacy sensitive issues? European provider does not automatically mean they aren’t logging everything or that they’d nor share that with a bidder?
No the OP, but I keep a $10 rolling budget on vast.ai. I tend to use it more for training things, but equally as useful. I pay about $0.15/hour for an A4000.
OTOH if you are really concerned about privacy it's probably not the best choice (it's random people's machines).
I want to propose a called AI Index, if you can do say 1000t/s on processing and 50t/s on generation on a min 25B model you can call your device AI9 and you have AI7 and AI5 and AI3 with 3 being 100t/s processing and 10t/s generation on a min 10B model.
Values above are on a 100k context size minimum.
OFFLINE.
So any effing claim of being an AI device doesn't end on the demo of best buy laptop where you remove a dog from a photo....
Or that you have "better noise removal" or "you look better on the cam" because that's called noise cancelling and instagram filter and it's being done for ages...
You can run a model at 1000t/s on a low and android phone. The output will be dogshit bad, but it'll pump tokens like nobody's business. The problem with what you're proposing is that there's no baseline definition of what those tokens are (for what purpose? of what quality?). And pinning that to a certain model or level of correctness means your benchmark becomes almost immediately obsolete.
That's exactly what I am doing. I have a Dell R740 with 96 cores and 512 GB of RAM, and RAID 10 with 16 SSDs. All of my machines connect to it via Tailscale.
Did the same. Took a week or so to get used to the workflow after doing dev locally my entire career, but now I'm certain it's the best way to work. Even better is using a few scripts to simplify things like opening in VS Code (instead of `code folder-name` I run `codeon repo-name` from my local machine).
I did actually get a new M5 Pro/48GB machine last month, but that's only because I bought my work machine to keep my personal and professional lives properly separated; otherwise I wasn't seeing any issue running that M3 Max/36GB for a long time further.
I've just transitioned to the same setup a week ago - that's why I'm considering Air as the next upgrade, given the better portability.
One thing that bugs me mentally is leaving that machine pending on my input ("wasted") overnight etc... I'd like to arrive at a workflow where I can have agents running toward the overall goal autonomously overnight as much as they can, and I can review and iterate and "engineer" parts that need it afterwards...
For now, I still guide implementation very closely, piece by piece. Has anyone had any good experiences with workflows / skills / similar for that?
I haven't had good outcomes from letting the agent decide what to do, but if you spend time deciding what tasks to add to your backlog, then how to accomplish those tasks, when you're awake, you can end up with five or six or more projects to let your agent do overnight. planning when you're awake, execution when you're not.
I have a Hetzner box that costs about 10$ a month - I use that as my always-on box. I use Pi as my harness and they run in herdr. I can then from my Mac (bottom of the line macbook air - next laptop will be a Neo) ssh to the box and control my agents. I can also ssh into the box from my iPhone (using Termius) and continue seamlessly from there too.
Have the same setup -- got a good black friday deal last year on an AMD 395+ w/ 128gb of RAM. Still not super bullish on running AI models on it, but it makes a great dev server anyway.
Actually had claude spin up a kanban board to manage my tasks in a 'TODO -> Plan -> Implement -> Review -> QA -> Shipping -> Done' flow (with a worktree for each task) that works great!
If that continues to be the case, next time you can "downgrade" to the MacBook Air. Even today you can get one with 32GB RAM, so when you're in the market I'm sure more will be available. They are quite thin and light compared to the MBPs.
Yep, that's why this M6 Pro and its ancestors probably exists. It's for people whose work can run on MacBook Air specs but with the superior screen, battery, speakers, and ports.
True. I still think it is good. My dream is for 120Hz on a future Air model. I took it as a positive sign they brought 120Hz to the non-Pro iPhones with the models last year. I can dream, I guess. And yes, I know MBP screen has benefits other than the refresh rate.
> all I need is a laptop with a big/nice enough screen to ssh
> 48GB 16 inch MBP
That's wicked overkill for ssh + browser. A 15" MacBook Air would work for this really nicely. This is how I use my 13" Air, and for the big screen I dock it at a desk.
I wish they could bring back the 11" Air or 12" MacBook and trim the bezels for something like this. It would be an amazingly portable computer. The 11" iPad Pro with keyboard case is almost this, but you're stuck with iOS and it's actually pretty thick and heavy compared to a MacBook.
>With the new 14-inch MacBook Pro, users can connect high-speed peripherals, transfer photos, work with multiple external displays, and more with its wide array of ports.
* looks inside *
4 kinds of ports. while being generous, and including the SD card slot and the headphone jack.
I don't know anyone who would consider 2-4 to be a "wide array".
It's a shame that only the Mac Studios come with USB-A ports. I know, won't fit on a MBA, but similar size and class laptops to the MBP often come with at least one USB-A.
It's just handy when you have a USB-A device and don't want a dongle, traveling, etc.
Am I crazy, or are a lot more people using the term "uplift" in place of "increase" within the last year? Google Trends (for whatever it's worth) shows a significant increase (uplift?) since mid-late 2025. [0]
Language changes. People at work have started saying things are an easy lift instead of a quick win. And it's not new hires, it's the people I've worked with for years. I never got on board with quick win either.
Some people adopt stuff faster than others. I've lived in NZ for nearly 20 years and called someone bro for the first time just last week while playing a sport "should have left it for me bro". Slipped out natural as. Still having bad thoughts about it.
It's a base M6 chip. The Pro doubles that, and the Max doubles again. And the Ultra doubles yet again. This is how the entire M series has always worked.
The M6 version is also cheaper than the M5 Max by a significant amount. I wouldn't be surprised if the M6 could keep up on LLM prefill (compute bound), but the memory bandwidth means that decode is going to be a lot slower, most likely?
If you want to do anything beyond getting a small single-core speed bump the M5 Pro is a better, more capable purchase. At least until the M7 comes out.
It's also absolutely wild that you apparently cannot spec one with more than 32GB, for any amount of money? Wow.
Not that I blame Apple here, for once. Seems like they're trying to cope with the limited quantity of RAM they can procure. And with the paltry bandwidth (170GB/s) of these CPUs they're a bad choice for anything but tiny local models anyway.
That's for the base model M6. The Pro can take more, the Max even more, the Ultra the most.
They're not making those models for the M6 though; they rerouted their resources to making M7 faster at running local LLMs.
If you want more bandwidth, get an M5 Pro, Max, or Ultra. If you want a lower power base model, get the M6. If you want to have your cake and eat it too, wait for the M7, reportedly in the spring.
Umm yes, the M6 chip only has a certain amount of memory controllers and channels, which means you can only have so many LPDDR5X modules.
Yes, high-capacity LPDDR5X are made and do exist, but for all practical intents and purposes, you can't buy enough of them to deliver to customers at Apple scale.
I thought we were supposed to get a design overhaul this gen?
Also framing AI use for the base M models marketing makes no sense to me. You are way better off getting a used M1 max with 64gb and way higher mem bandwidth when those are the real bottlenecks.
Apparently another announcement coming tomorrow, with the OLED panels. But according to the fpt channel on Youtube, it doesn't look any different. Just has an OLED panel with touchscreen, with massive cooling changes inside.
MacRumors is still claiming we're getting a new chassis
> Apple on Wednesday will announce 14-inch and 16-inch MacBook Pro models featuring touch-enabled OLED displays with thinner bezels, a Dynamic Island instead of a notch, M5 Pro and M5 Max chips, and a lighter and slightly thinner design
Interesting. Will be way out of my budget, but keen to see how the OLED display performs long term. Nearly everyone I've seen that has used an OLED on laptop say they absolutely love it. But I still hate the idea of touchscreen, especially with the changes to MacOS to cater for this
IMO changes to a touchscreen would've been perfect for professional artists, but only if it enabled using an Apple Pencil. Which is pretty unlikely if they don't also solve the ergonomics of doing that, e.g. with a 360" hinge.
Look at the first image with the people in the field.
That looks to me like a very poorly generated image, not a real image of people, and not even a high-quality AI generated image.
Is this what we can expect of the new apple?
I'm not against AI generated imagery, but I expect them to be of equal quality to a real photo.
Or am I wrong, and it isn't AI generated?
The things that make me think it is just poorly made are
1) the angles of the laptop on the black case,
2) the lack of any lighting on the back panel,
3) the size of the hand on the woman in the back, and the squareness of what she is holding (if it is even supposed to be another mac).
That was my first thought, is that guy's face AI? The angle of the laptop looks off. Then I decided I don't care and moved on, but yeah, you're not the only one who thought it looked off.
Almost certainly not AI gen, it's shot with a 35mm lens (quality looks like the 1.4 L to me), probably around f8 given how the DOF falloff happens, heavy forward flood (flash or otherwise, but whatever it was it's defused) and then what's I think giving you the sense it's AI is it looks like they've over used a double S curve to bump the contrast and sharpness. https://www.youtube.com/watch?v=XZh_9HnO_4w&t=640s
It is interesting how people are confusing depth of field with AI generated. It's a shame, but perhaps it will end the bokeh conversations and disparaging of M43 sensors.
This is a strange release. The M6 tops out at 32GB RAM and then you fall back to M5/Pro. Kinda midway between an Air and a true Pro. Only 14" no 16" with M6.
Hmm, still no OLED displays? I wonder what's the holdup? (I have a fairly mainstream PC laptop with OLED display, and the display is the one thing where my MBP looks pretty poor in direct comparison).
I just bought a framework 13 pro, going to main that into the ground, 15 years+, with maybe 1 allowed motherboard upgrade in that timeframe, but hoping to completely avoid it and just do repairs and minor upgrades.
Just now realizing there's a difference between "MacBook Pro with M# chip" and "MacBook Pro with M# Pro chip". Was curious to see how my MBP with M4 Pro compared, and I saw it has 273 gb/s memory bandwidth compared to the MBP M6 (sans Pro)'s 170gb/s. Makes sense that "Pro" in the laptop model doesn't also describe the chip, but only if you remember that the non-pro "Macbook" line stopped in 2017 (which was the last time I went shopping for Apple laptops until buying the MBP4)
The "Pro" in MacBook Pro has slowly degraded to "it comes with a fan so that the chip can keep chugging when it gets a hard workload, unlike the MacBook Air where the chip will throttle since there's no fan".
If anything, Pro really means pro now. It used to just be the default laptop everyone bought because the Air was too slow and had a terrible screen. The 2020 MBP was almost as thin as the MBA and had barely any more connectivity. Now it's bulky and provides more power than normal users need.
I have an M1 Air 13" and M2 Air 15", and M4 Air 14" and and M4 Pro 14" with nano texture.
I love them all. The M4 Air 14" is from work, so kinda limited in what I can do. The M4 14" Pro is kinda chunky after the Airs. The M2 15" is ideal except on a plane. The M1 13" is great as a personal device on work travel. If only I could not work...
As lots of Apple Users are in here: I'm thinking about "sidegrading", my M2 Pro runs fine, but I chose 16 GB/512 GB when I bought it and I think 32 GB/1 TB would be nicer in some cases.
Is it easily possible to sidegrade/upgrade with a Timemachine backup?
I checked used prices though, and I think the market is not in our favor. rebuy.de has an M2 Max with 32 GB/1 TB configuration in "very good" condition and is asking 2100 Euro for it.
I've been pretty happy with my M1-2 devices like most posters here seem to say, but I don't have the same level of core stability while doing dev work. Am I doing something wrong with my workloads?
- Orbstack, but just for one containerised service
- Laravel Herd for app environments, so all native services instead of containers with bigger VM overhead
- Brave browser, profile for work and personal
- PHPStorm (quite a hog on RAM)
- Now agents in the mix
Had an M1 Pro 8GB but I found 8 to really not be enough. M2 Air 16GB has been a much better sweet spot, albeit it feels like the battery life on it is worse than M1. I've recently started trying to use Zed more and more as my editor instead of PHPStorm and things feel mostly stable, but I still feel like a beefier machine would be good for work purposes.
Years ago there were frequent crashes (kernel panics) i could directly relate to IntelliJ (PHPStorm, etc), so it might sound dull but upgrading to the latest versions is the first thing I’d check. In this case even a manual check on their website instead of relying on the auto updater to work.
I don't think you're doing anything wrong. I struggle with similar workloads on 32GB RAM. Compute isn't the issue, once I have a JetBrains IDE open, a dev environment, some browser tabs, Claude, and a bunch of other crap Electron apps I need the machine is already swapping quite a bit.
I have found that my M5pro gets unstable when I really stress it's memory use. I think that with 8 or 16GB you are going to hit this quite a lot more than I do with 48GB.
My big issue with the M1 generation is the lack of more E-cores, which get assigned a lot of the OS background tasks. When working at the top of the RAM limit, I'd notice issues around lag to the point that I started looking at CPU stats and found those two measly e-cores completely maxed out for hours on end. Occasionally a crash, though that could be RAM related rather than CPU.
A bit of googling around seems to point that architecture as somewhat of a problem for others, and might be why Apple quickly moved to put more e-cores on the later generations.
Something to consider. Never managed to get proper distinct confirmation that this was the problem, but a (corporate) laptop upgrade did get me out of this pickle.
The text of this product's announcement really tries to sell it as a 10x AI performance upgrade for existing MacBook M1 users.
Ok, plausible number for that use case. Striking how it seems like the M-series Macs are almost too good enough for most people.
It's almost desperate prose.
(The AI image processing in Pixelmator Pro saved me hours of work on a recent project. Two button clicks vs a day of tweaking. But even that was on my M1 MacBook Air.)
Is this an odd rollout for their actual high end MacBook Pros, or is this sitting in that space where they like to have a base MacBook Pro between their Airs and the expensive MacBook Extremely Pros?
Just the standard chip without Pro/Max makes me assume this is that base one. If so, the price is way high for where that tends to sit, but welcome to buying a computer in 2026 I guess.
This is the base one. You get a better display, more, faster Thunderbolt ports, and an honest to god HDMI port, plus some fans for cooling on long running processes that you don't get with the M6 Air (which also doesn't exist).
I have to say, I really dislike the thing where the base M(n+1) is always in the market alongside M(n) Pro/Max. I appreciate the underlying chips are probably quite distinct, but from a naming standpoint its confusing as hell.
makes me want to wait another one
I have an M1 Max and just set up playing with an M1 (non-max) and realized it can't drive 3 displays. I recall getting Max just for this, but generally, makes me think of non-Max as a non-pro after all.
And having what's presumably the latest CPU but with potentially worse features than the last generation combined with increased prices from RAM is a deal breaker so I'll guess I'll push my M1 for another year.
Whenever I do this, I enter into a perpetual one-more-year wait loop.
When I finally cave in and buy something after 5 years I’m very momentarily sad for not perfectly timing the release window, then I realize I keep hardware long enough a half year barely matters.
I get you, but I suspect it reflects reality. The M-with-adjective chips are M(n) chips with more cores and cache, it stands to reason that it takes real wall time for TSMC to start cranking them out with acceptable defect rates.
So Apple can either toss the early production in inventory while the whole process catches up, or they can do what they do. Better for both the company and the consumer that they do the latter.
I'm guessing the Base and Pro chips come off the same wafer, and they get 5 Base chips (some defects) for every one Pro chip (very few defects). So they run the Base-model assembly line for six months and eventually they accumulate enough Pro chips to satisfy Pro-model demand.
You mean as in binning? High defect rates for chips but they're burned to disable non-working parts and sell them as non-pro chips and accumulate the good ones ?
The Pro and Max chips have a different die than the base model. They don't bin down to the base model. Maybe Max bins down to Pro now that those are the same die, or Pro bins up to one half a Max?
Pro chips come much later than base chips and vary significantly in ways other than number of cores (which can be up to double), such as doubled memory bandwidth and peripheral support (thunderbolt 5 instead of 4).
If you look at the platform name and the kernel code you'l realize that this is not even remotely true. The base M chips are T8xxx which are (more or less) just iPad chips. The pro/max/ultra are T6xxx and are their own new thing, starting with M1 pro which was T6000
I suspect it has depended a bit on whether a new process has lower output or higher defects. High defects, they'll focus on smaller baseline chips to get better yields. Low output, they'll launch pro models first because those are lower volume. Neither problem, they'll roll out updates pretty quickly across multiple product lines.
And that should change going forward with smaller chiplets being integrated in package, vs huge Ultra chips.
If the difference between M6 and M5 Pro or M5 Max is confusing to you, I strongly recommend you don't look at the names of literally any other CPU, GPU, or chipset that is being manufactured.
Not better - but Apple in early days was easy: Choose Air if you don't nedd much. Choose MacBook if you need a little but more and choose Pro if you need max power. That's one of the reasons why I bought an Apple years ago.
Compare that with the regilar lineup of other manufacturers: Cryptic model names which are uncomparable and models that are so similar that they could be equivalent.
So, no, 1 is not better but I don't care (as customer) about variety.
Well that was sort of the case. You had different years of each model to consider as well, which isn’t all that different from the M/Max/Pro thing we’re discussing here. They didn’t refresh all products the same year.
I think the Apple experience was better when they focused on delivering a single SKU of each product line, yes. Less decisions as to what to buy, greater focus on making that one SKU stand out.
Most of those models were not sold concurrently. The Apple II, II+, and IIe each replaced the previous model; the same was true of the IIc and IIc Plus. "//e" was an alternate textual rendering of IIe, not a separate model.
The most Apple II models that were sold concurrently were the IIe Platinum, IIc Plus, and IIgs. (And this was around 1990, at which point there were far more models of Macintosh computers on the market.)
The second one was the first iPhone with 3G networking (which had broadly been considered a required feature for smartphones when the original iPhone was introduced without it.)
And the iPhone XR! All part of the twelfth-gen models. Of course the subsequent thirteenth-gen models were just called "iPhone 11" in a return to sanity.
Tell me you're not old enough to remember the Performa and Centris lines without telling me you're not old enough to remember the Performa and Centris lines.
All the people under you not understanding that you meant AMD repackaging 2-4 year old CPU cores as """new""" by giving them a current series name and bolting some extra features on.
Put another way, it is a peculiar naming regime when M(n-[2-4]) Pro/Max/Ultra is superior to M(n). This is very visible right now because of the demand for running local AI on Macs.
It's also kinda sad that there's no clear-cut best chip, e.g. the M6 outperforms the M5 Pro in some ways (singlethreaded performance, maybe NPU and stuff) but then the M5 Pro also outperforms the M6 in some ways (multithreading, graphics). This was even worse when the Mac Studio was stuck on M3 Ultra for a while when the M5 Max was out. The M5 Max was simply better at most things and it was ironic that the fastest computer you could buy was briefly a laptop.
How could it possibly be otherwise? Single core performance goes up every year, but not enough that this year’s 12-core CPU beats last year’s 18-core CPU at multithreaded benchmarks.
It's also not exclusive to Apple; the normal product cadence from Intel and AMD usually has the large workstation and server chips falling behind the consumer chips in peak single-core performance for part of the product cycle. Some market segments just don't sustain annual updates very well, and it's always easier to get a smaller chip out the door first before a larger chip.
You can just Google it. It's not a complex answer:
- Snapdragon 8 is the highest-end mobile SOC line, as opposed to midrange Snapdragon 7 and low-end Snapdragon 4.
- Elite means that the CPU core is Snapdragon Oryon, and not Arm Cortex off-the-shelf designs.
- Extreme is a trim/upgrade package that gives the GPU more power and performance headroom.
- Gen 6 tells you the generation of chip from that line.
You should refer to prior discussions on this topic if you're genuinely confused, or somehow think that Apple has a superior taxonomy: https://news.ycombinator.com/item?id=50116504
Latest vs. greatest. Buyers who’ll pay for performance can deal with that discomfort and be squeezed for buying the best, but those who need an extra push to buy Apple might be sold on the idea of the latest.
I would guess there's still quite a lot of people on Intel Macs out there (they were sold 5 years ago), so it is still very much worth it for Apple to market to them.
Why? Apple devices have the best battery life in the market right now to my knowledge. It's an important part of the evaluation for a laptop which is the opposite of cringe.
Actually, they don’t! Intel Core Ultra Series 3 beats Apple laptops as of the M5 on battery life.
IIRC the Framework 13 Pro beats the MacBook Pro at both video streaming tests and office productivity with 300 bits brightness tests. (Framework live-streamed the video playback test, and the office provided productivity tests I referred to were done by Just Josh Tech)
Perhaps the M6 has leapfrogged but that remains to be seen.
In usual developer type tasks, among my friend group what I hear is that panther lake laptops are competitive with M4 laptops. But, M4s do that with a 55Wh battery, whereas most panther lake laptops come with a 70Wh battery. So if you're discussing chip performance per watt then it's probably an edge to apple silicon.
I did. I am seeing single core geekbench scores of 2700 compared to my MacBook Air at 3500. This, for a laptop that is thicker and heavier than my air.
in my original post, i said "perf", i am not sure what you imagined, but i am not moving anything, geekbench single core score is a perfectly reasonable thing to measure under the name "perf"
1. It’s Geekbench, not a real experience of using the system. A geekbench number being xx% slower doesn’t mean that the system is xx% slower at doing tasks. It’s a synthetic benchmark.
2. The Intel chips beat out the M5 in multi-core and GPU.
3. These Geekbench scores are close enough that they are in the same performance class. We aren’t talking about 1,000 versus 3,500. The end user isn’t noticing a performance difference between 2700 and 3500, especially since that number is single core only (most workloads can take advantage of multiple threads these days).
Even if we concede that the Intel chip is slower for single core tasks only we can’t just lie and say it’s not in the same performance class. It’s very clearly comparable by objective data.
What we can say definitively by objective measures is that Intel is behind by a year at most, which essentially doesn’t matter in the grand scheme of “which laptop to I buy?” It’s close enough that the processor shouldn’t be part of the buying consideration for the majority of consumers. Intel is not years and years behind what Apple is putting out there, this is not the M1 where Apple was beating Intel in performance and getting double the battery life. We are looking at much more incremental differences where the buyer will not notice a difference over the lifespan of the product.
multi core scores? So you'll happily accept a 100,000-core pentium one class CPU to replace your laptop? it will have higher multicore score. No? Really? Oh? why not?
I recall seeing some crazy stats like 41+ hours maybe on a snapdragon processor w/windows. I imagine your screen brightness would have to be like 99% dark and not playing videos is my guess.
There's no happy medium here that would satisfy everybody: when they make marketing claims about improvements relative to products from several years ago, they get criticized for not focusing on the improvements relative to last gen (even if those comparisons are also somewhere in the announcement). When they focus on the most recent generations, people complain that it doesn't make sense to upgrade every year.
Just accept that even if you aren't still using an Intel MacBook Pro, the comparison is still relevant to some part of their target market.
They only just dropped support for Intel Macs from this fall's operating system release (but are still doing security patches for the releases that Intel Macs can run). A year from now, it actually will be fairly reasonable to expect them to give up marketing to the holdouts.
Just Josh Tech’s review of the Framework 13 Pro or the Lenovo ThinkPad X9 15p Aura Edition. Multiple similar sized/classed laptops get better battery life.
This is of course as of the M5, M6 hasn’t been tested yet.
last time I checked, every Windows laptop had the absolutely massive caveat that it needed to throttle performance by a ton to have a low enough TDP to compete on battery life. You just cannot match the efficiency of ARM Apple chips with an x86 chip.
As far as I can tell after just looking at the specs, what I said is still the same. For example, an Intel Core Ultra 5 Processor 338H has a TDP of 25-65W vs. the M5 TDP of 3-27W. These aren't even directly comparable, but it gets worse going higher in the Intel Stack.
Further, every Windows laptop seems to have a worse display. For example, an M5 Macbook Pro 14" has a 120Hz 254ppi mini-LED display with a peak brightness of 1600 nits vs. a Dell XPS 14" with a 120Hz 161ppi LCD display with a peak brightness of 500 nits.
So it seems like they have figured out that they can put bigger batteries, worse displays, and throttle TDP to get better battery life, but it is still not possible for them to beat Apple's efficiency just based on those TDP numbers alone—these chips are just drawing more than 2x the power at peak load.
I don't understand why there isn't more effort towards ARM in the Windows world.
I think it makes sense. Do you think most potential buyers are coming from last year's model, or from laptops that are more like 3-5 years old?
As long as they're clear about the comparisons they're making ("X% more battery life than Y" as opposed to simply ("X% more battery life") I'm good with it.
I’ll wait for it to surface on eBay in few years and then consider again. Pretty sure many of these go in the hands of people who can never make good use of all the nice hardware in it.
I am due to an upgrade (still using the 2019 intel). However I am torn between a M5 Pro with 48 Gigs of ram or waiting for M6 pro. I’d really like to run local models for once.
Interesting. I guess I’ll have to go with the m5. On one hand it feels bad to buy older generation hardware. But then. It’s already so much faster than what I have…
Still using my 8gig m1 to make mac builds, its slow, but works fine. Keep thinking about upgrading, but max 32gigs ram, if I can hold out for 2 years, and maybe have 64/128gig as options, and cheaper.
In your shoes, I'd just get whatever Pro processor is available at that time. (Knowing that new processors were coming, I happily bought an M5 Pro machine last month)
Why would you want greasy fingerprints all over your display and gorilla arm syndrome when there's a perfectly fine keyboard and touchpad attached to the display for free?
No reason for gorilla-arm when touch is an alternative interaction mode and not a required one. You still have trackpad and keyboard. You can dynamically switch modes depending on the action.
Gorilla-arm was a fear based on the naive idea that if you had touch, you would be required to only use touch for every action.
Is apple going to be so restrained to keep the design of the OS as a mouse first experience or are they going to cripple information density so I might be able to close a window with my elbow?
People act as if Chromebooks with touch screens aren't already considered a huge UX upgrade over Chromebooks without. Use your finger where it makes sense and the mouse where it doesn't. Literally nobody gets tripped up over this in practice.
Chromebooks run Android apps natively, so you can flip the screen around and use them as pure tablets or just use tablet apps in laptop mode just fine. For math, you can just write directly on the screen. In schools, kids playing Blooket/Gimkit/Kahoot on touchscreen devices crush those using a touchpad.
I want the newest one and I want to run Gentoo Linux on it with working Asahi patches for all aspects of the hardware. I want it enough I'm willing to pay for it. I'll pay 3x Apple's current sticker price.
But Apple will never sell that to me, as far as I can guess. Apple is not a good person.
15 U.S. Code § 7 - “Person” or “persons” defined
The word “person”, or “persons”, wherever used in sections 1 to 7 of this title shall be deemed to include corporations
https://www.law.cornell.edu/uscode/text/15/7
Still using the M1 Macbook Air for 4K video editing on Davinci Resolve and Photo Editing with Darktable (open source). Just 8GB of Ram too and I've never really needed more. Obviously these new versions are great but even the original is not too shabby.
The design is really great for the new MacBook Pro. The cpu is no compare to other laptops.
One thing, framing AI use for the base M models marketing makes no sense to me.
But honestly this computer is still faster than any of my laptops that came before, even with the visual atrocity of Liquid Glass.
I'm currently having an M2, but I just can't imagine buying an Apple laptop again. The hardware is great, but the sofware is just unbearably bad. I used Arch Linux with KDE before, and it was just an experience 10 times better. If they would just support Linux out of the box...
I would not get this because I want to run LLM's locally, you are really gonna get gpt 4 performance and constantly looking at the SOTA or the open models that needs way more to have it work practically, and without your laptop sounding and feel like a hair dryer.
I need 64GB RAM and 1TB SSD suddenly having been more than comfortable with 16GB and 256GB respectively. My agents have sessions which have workteees and scratchpads and simulators and build runs and test suites and wow. The hours I’ve been spending for memory swaps and quad simulator calls to run through. I am compelled to view a $4k machine as a necessity. Also a cool thing about M5 Pro is that it’s very close to Max on the CPU performance side with identical core count if you get the 18 core variant. The Pro’s Geekbench scores are also in similar ballpark as the Max. You rather get the Pro for most use cases and use the extra cash on RAM.
One of the interesting slights of hand Apple marketing must do is simultaneously say the new processor is awesome while selling the higher spec versions of the previous processor as upgrades:
The M6 chip joins M5 Pro and M5 Max to create our most advanced line-up of chips ever built for a pro laptop. With a 12-core CPU, 12-core GPU and our first ever Dual 16‑core Neural Engine, M6 brings next-generation speed and advanced on device AI. All three chips deliver phenomenal single- and multithreaded CPU performance and faster unified memory. . . .
The case of your 5090 having become a coveted artifact of great rarity in these times - may be a nuance of some measure.
Also, the following two have very different 'needs' to feel good enough:
1. heavy agentic workloads via harness
vs
2. a simple private chat interface with a local instance
Even if an MBP's bandwidth were higher, it's tough to see a reason to run it as a constantly loaded 24/7 inference machine due to heat degradation.
P.S. still extremely awesome to be able to host and chat with a decently quantized 27B model on your laptop. ...and to be fair even a 32GB Air would do that!
It is a thing of beauty: the amount of folks this may add to the hacker/tinker community.
The 5090 does have a bandwidth advantage, but also must pass more data from main to GPU memory for any particular task. Comparisons of a 5090 to a DGX Spark at 273 GB/s are illustrative since the Spark uses unified memory similar to Apple's M-series. [0] The 5090 certainly wins most benchmarks but that doesn't mean the DGX Spark is not useful.
Loading data to the GPU will generally never be the bottleneck. Yes that's a bit better for unified memory, but prefill and decode will still be bound by compute and memory bandwidth respectively.
There's some value in having a huge amount of memory addressable by the GPU for loading larger models, but even that doesn't make sense for current gen models. For MOE models, newer inference engines with smarter GPU expert caching and pinned memory allow a GPU with lots of bandwidth to significantly outperform a Spark, so that leaves you with larger dense models that fit on a spark but not on a GPU where you might have a use-case.
But even then it's hard. For instance a full-fat BF16 of a ~30B parameter model will fit on a 128GB spark and not a 32GB 5090, but at that size it's going to be frustratingly slow on the spark given the compute and bandwidth constraints.
Imo these devices are mostly good for experimentation, or if you only want to do overnight jobs or training which wouldn't be possible on a GPU, and you don't want to pay $100k for a DGX station.
About time they put a battery in the lid, which is what an iPad already manages to do with thinness. Boom another half a decade of updates all the way to 48hrs
It's confusingly worded but I think it's a suggestion for something Apple could do in the future. I don't see any mention of this in the linked Apple article.
I'm currently on a marathon to keep my M1 Pro running until the M7 Pro/Max is available. Not sure if I'll make it, but I'm not eager to pay $6000+ for a decent spec dev computer, until I really have to
My M1 Pro has been great, certainly doesn’t feel like a 5-year old computer.
M1 Air guy here... I'm not looking for a replacement still. Almost 6 years now. Runs butter-smooth. Best tech purchase I ever made.
Have an M1 air as well. Got a Pro m5. For daily coding tasks, and general web browsing the M1 air still holds. I'm going to pass it on to my kid. Reminds me of the honda we have that we'll also pass down.
The M series MBAirs, and ofc the M1 originally, are hands down THE BEST laptops ever made. I have a maxed out MBP ofc but the Airs are unbeatable quality and price/performance.
I think the M1+ MacBook Airs will be regarded eventually like the original Macintosh originally the iPod, or the introduction of the original Air and the creation of the product class that came to be known as “ultrabooks”. It really changed everything about laptops to have a fantastic keyboard, speakers, screen, trackpad, hinge, and 12+ hour battery life… for $1k.
Still running my M1 Pro as well, no plans on upgrading for the next few years
Biggest issue with my M3 Max is apple mandatory fun bloatware. I think it's the same playbook as phones long term. The hardware is still powerful but the software is optimized for the latest CPU, so older models run hot, take a fraction of the second longer to load the resources, etc. That stacks up over 5-6 years of releases to make older models feel slower.
I've held back an OS update or two, specifically because I'm worried about performance dropoff on the old chip
But with all of the AI-discovered exploits I think it'll become more important to keep everything up to date
Apple makes security updates available for the last few major OS versions, and even backports really bad ones further back on a case by case basis.
Same, and I dual-boot Asahi as well.
generally agree, although mine's battery health has been sitting at 80% for what seems like ages now.
I’ve been a Mac user for 20 years and it’s wild to think how slow my 2003 PowerBook G4 or even my 2008 MacBook Pro felt after five years compared with my 2021 M1 Pro.
My 2013 Core i7 took almost eight years to get to where the previous ones were at five, and if it weren’t for local LLMs pushing me toward an upgrade in the next year or two, I have to imagine my M1 Pro could be my daily driver well into the next decade.
My 16" M1 Pro personal laptop is literally just as smooth for 99% of tasks as my work M5 pro. It's crazy how well they built these. I mean the screen on this and the speakers are superb. And you can find these for like 500-800 bucks 2nd hand. I just hope they give us a few more years of MacOS updates.
My work gave me an m4 MacBook Pro. I could barely tell the difference between my personal computer and my work computer.
I'm wondering what decrease in performance you were expecting from a five year old computer? It isn't like the processor suddenly gets slower simply because it's older.
I suspect if you used an average 5 year old PC and current PC, there'd be very little noticeable difference between the two outside of LLMs
I rode a 2009 MacPro5,1 (¡galliently! into the future, with capable GPU upgrades) until Jan23's M2Pro mini debut (which I still use as my main Mac).
Apple Silicon is certainly incredible (I also have a base-model M4mini, but it severely lacks RAM/ports so I don't really use it much/eWaste); my only Silicon-purchase-regret was buying a beautiful new 2024 MacBookAir (15") – heavily discounted $850@CostCo – but with only 8GB of RAM (it is functionally worthless, other than for watching it destroy your SSD with page-outs as you casually stream YouTubies...). In 2024, only 8GB of RAM (soldered onto the board because fuck you)!
Macbook Pro 14" Max M1 (64GB). Best Apple device I've owned and worth the investment at the time, especially given the prices now. It's still pretty great for being as old as it is. Have no intention to pay inflated prices, I'll ride it out.
If the rumours are true (about AI capabilities) it’d be best to wait yes.
which rumours?
The M7 will the much better for local LLM inference. If you can afford the memory.
Oh interesting, any more info on "how"?
With a three point plan, I suspect: memory bandwidth, memory bandwidth, and more memory bandwidth.
How useful would that actually be for local inference, when the bottleneck is, essentially, always memory size? I think the sweet spot is somewhere around 64 to 128 GB. That ought to be enough for almost all uses, like Qwen3.8-Flash-Next at a decent quant.
You can order a 128GB MacBook Pro today -- though point taken, memory size probably does belong in the three part plan :-)
> I think the sweet spot is somewhere around 64 to 128 GB. That ought to be enough for almost all uses
This will be the “640kb ought to be enough for anybody” Bill Gates quote of our era.
Memory size is a bottleneck, but I have found that speed is very important for getting actual work done. Running a big model is great, but if you're getting 10t/s you're going to be spending a lot of time waiting.
More ASIC Mat-Mul too.
Plus with more low precision compute… to help with the memory bandwidth!
I'm hodl with my 2018 Intel mbp for a truly local model capable machine.
I'm loving my M1 Macbook Air still, and only a few weeks away from six years of use Sometimes it lags a bit when I'm running a macOS VM, and all the coding agent VMs, so I'm kind of interested in going to 32GB of RAM, but honestly this computer is still faster than any of my laptops that came before, even with the visual atrocity of Liquid Glass.
I'll go and look at this new box, drool a little, and realize that I'd get more personal utility from donating $3k (or whatever the trade in delta is) to a charity than upgrading my computer.
I freaking hate liquid-glass UI Apple has "forced" onto us. No freaking way to disable it. I hate it so much.
There's a slider to reduce the liquidity
Reduce != disable.
Big Tech found the wrong slider
Been on HN for many years now, and the one thing that never changes is people posting angrily about small UI changes. It makes me wonder- if things have only ever gotten worse, what was the peak UI? MS DOS?
Small improvements don't get much accolades. And for the most part, the UX is so good that most changes tend to be neutral or for the worse. And when you're really attached to something and there's on an unnecessary change, it feels almost like betrayal!
I, for one, prefer Apple products because they sweat the small stuff more than the other options, and when they have put in effort to make things worse it's kind of infuriating.
I do a bit of desktop Linux work too, and it's OK, but hard to complain when it's all open source volunteer work! Windows is so terrible that I wouldn't even know where to start. What's the point? There's nobody at Microsoft that even gives a damn.
No, it was Windows 3.11
Windows 2000 / XP in classic mode / XP / Vista / 7. For OSX I'd say High Sierra to Catalina.
I’d go with Tiger (10.4) for peak OS X, personally.
Leopard was better because it introduced spaces. Honestly I think that whole concept (I know windows has something analogous now) was the biggest step change in OS functionality I've seen in my life having started on windows 95.
I never used Spaces much.
Apple was actually last to the party on that. Linux had virtual desktops for years at that point, and Windows had them as well through Microsoft PowerToys. I’m pretty sure I was running it on Windows 2000.
I had virtual desktops on UNIXes with X11 well before Linus started working on Linux. They came for free to Linux because they were implemented by many window managers.
Not Linux, the X Window System. The first implementation in a window manager predates Linux by a couple of years. And to no surprise Xerox had them first
The problem is that window management is so atrocious in Mac OS that using multiple desktops is just asking for trouble. Like of you close one window of some software, it will jump to some random other window of the same software on some other desktop.
In the end you’ll spend a significant amount of your time trying to “get back” to where you just where (that alt-tab, a possible savior for this in all other OSs, doesn’t work at all on Mac just adds to the UX mess)
Ehh I'm not to caught up with window management on macos. Bunch of shells open I might use tmux. But otherwise I just keep things in piles organized in spaces, and use expose all to get to whatever I need in the pile. It works much like the piles on my desk in real life so I'm comfortable with it. But it was better in leopard when it was a grid of spaces vs a line of spaces. Less rigidity was nice.
But even then. If you close a terminal, it'll jump to some random terminal on another desktop. It just doesn't make any sense.
It doesn't make a lot of sense to me as a default, but there's an option to turn off the automatic switching of spaces when switching applications, and it pretty much eliminates the unwanted switches from closing the last window on the current space.
The separation of Cmd-tab and Cmd-backtick is one of my favorite things in Appleland... what about it is broken to you compared to alt-tab?
Eh. I get what it's going for. The Multi-Window application concept and what not.
In the end though I just feel like it's extra cognitive load to have to think about two different shortcuts to do what is in many ways the same thing.
Say, I've got two browser windows open and a code editor open. I'm wanting to flip through the browser windows to check on things while I work in the code editor. In reality I'm just wanting to juggle the three windows, but depending on if I'm wanting to go from browser to browser it's a different shortcut than if I'm wanting to go from editor to browser. And there's a good chance I'll then need to swap from editor to browser, then issue a different shortcut to go from that browser to the other. When I could have just hit the same shortcut twice.
I've just never really encountered a situation where I've felt like it added something, but I'm constantly having cases where its more annoying.
Thanks, that does make sense. Though as a many-windower alt-tab is just often tooooooo big for my tastes to even try to rely on.
The 40-year-deep-history Mac answer to workflows like that with a small working set with frequent cross-app-one-to-other-switching IMO is mouse-based navigation + spatial partially-overlapping windows, but this seems to be an extremely foreign concept to Windows/*nix users that grew up on full-screen-maximize or tiling. So I don't think is particularly likely to be adopted by anyone who doesn't already have it burned into their brains at this point. :) There's a lot of these old Mac bones still baked into the OS (e.g. if you have a folder on your desktop, and double click it, and move it, then close it, it'll open to the location you moved it to next time. And that memory is separate from that for another folder on your desktop. Though some of the follow-on behaviors have changed (opening folders in already-open windows used to behave similarly, opening new windows by default).
The 20-year-history version is my own personal lord and savior: Exposé hot corners.
Spaces and no-mouse navigation are still in many ways comparative afterthoughts to mouse-based/spatial workflows. Though as someone with 30+ year of Mac muscle memories in my head, the switching-windows-vs-switching-app concept and awareness of "where" I am relative to it is pretty embedded.
Firstly backtick doesn't go to the most recently used.
The whole window arranging so you can click on them doesn't help you when you have multiple desktops, close some terminal, and jump to some other terminal three desktops away. Like, arranging windows so you see some part of it always, just so you can click on it to switch back and forth between windows feels like a horrible workaround for broken alt-tab behavior. But there's additional levels of weirdnesses that just make using Apple frustrating and scary, like it's constantly throwing random curve balls. It just reeks disorganization. Which is really unfortunate - 10 years ago or so (when witcher task switcher still worked) I really enjoyed working mac. Back then I was mostly coding in the same windows I guess. Nowadays it's business work, and it's just a pain.
For me it is that cmd+` only cycles between windows of the same app in the active desktop. It makes it hard to switch back to a window of the app that was put in full screen.
I get that that behaviour shouldn’t just be changed but I would appreciate it if there were another key combination that would allow cycling through all windows of an app.
Expose added in Panther was the last window management feature that changed my day to day usage of PCs.
Expose + Cmd-tab/Cmd-backtick being separated == a gap that Linux/Windows have somehow never managed to close in 20+ years now.
Liquid Glass was not a "small UI change". It disrupted many accessibility aspects of macOS, so much so that Apple toned down quite a few aspects of the design in subsequent updates, even adding a slider to reduce the effect. [1]
The quantity of requests for how to disable Liquid Glass on Apple Discussions is staggering. [2]
[1]https://www.macrumors.com/2026/06/09/macos-golden-gate-liqui...
[2]https://discussions.apple.com/thread/256227663
Yea, I am trichromatic color blind, and it is legitimately unusable for me. I am unable to type on the keyboard. I know there are changes since it launched, by after my ipad updated automatically, rendering it near useless to me, I just disabled updates on my iphone.
I think the biggest thing isn't really whatever it is but the act of changing and forcing everyone to change into the new ui that is the disruptive part. It is a combination of window dressing changing but also sometimes functions breaking. It is ironic because software written for developers prides itself generally on maintaining compatibility with say depreciated flags, or if things will be depreciated plenty of notice ahead of time. But that only holds for cli software. GUI software and the OS that all goes out the window. Stuff breaks, workflows break, you have no recourse and in fact they might make it genuinely difficult to go back to an older version. Like this part of it is the shitty part. Never mind how it happens to look. It is the forcing effect.
I think it’s also because these UI changes are always designed to improve first-impression and first-usage. They are outright hostile to anyone who wants to go fast and get faster — the power user. At minimum the re-arrangement resets any gains you had, and at maximum its irrecoverable because the UI is now fundamentally designed to slow you down.
This eventually becomes awkwardly precise complaints about information density and general change; but the underlying issue is I want to go fast and the system says no
I think the only real solution to this is to switch to an OS with a backing org that has optimization for power users as a first-class goal. I've been liking Omarchy and NixOS for that. I say this as a Mac user since 1989.
I'd say Snow Leopard was peak.
Small?
Mac OS 7.6.1 was peak Apple
It was great if you could tolerate an application crash taking down the whole operating system hard.
It's been a long time, but I'm pretty sure Appearance Manager and the corresponding unreleased-but-leaked Hi-Tech theme was 8.5-or-so-and-up-only, which means I must reject 7.6 on aesthetic grounds ;)
On the Windows side, Windows 2000 and XP/Vista/7’s Classic mode, though Aero was quite nice. I’m not a fan of the interfaces from 8 onward, and I didn’t like Luna in XP.
On the Mac side, it’s a toss up between the platinum interface of Mac OS 8/9 and Mac OS X Tiger. With that said, I have a work-issued MacBook Pro running Golden Gate as my daily driver, and Liquid Glass is alright. It’s not what I’d prefer if I had a choice, but it’s not a dramatic departure from previous versions of macOS. A Cheetah or Puma Mac OS X user should be able to navigate Golden Gate with ease, mostly.
I am a HUGE sucker for Mac os8/9. I currently have a PPC Mac Mini and a Powerbook G4 that I tool around with occasionally for nostalgia's sake. I used slew of various G3 computers running OS9 as my main personal computers picked from the scrap bin at my dad's job (graphic imaging company) from ~2004 until I went off to college in 2010. I had some kind of modified Mozilla browser (Classilla I think?) that had modern-ish web support.
I've been on the lookout for a good condition G3 beige desktop, that was my first ever computer that was "mine" and I really want one just for the sentimental value.
If you consider the transition from macOS 15 to 26+ 'small', I think you should see an optician.
Like GP, I despise the new design. I'll say the old design was pretty peak design. Windows looked like windows. Buttons like buttons. With accessibility tweaks (for transparency and borders) I could make it very, very usable.
This new thing is a crime against UI. Every icon now sits in a circle, and takes up % more room as a result. Because … it's pretty? It isn't pretty.
I upgraded my laptop thinking I was going to develop an app and that I'd need to be on the latest design. I ended up developing a TUI so I needn't have bothered. Every time I see my partner's laptop, still on Sequoia, I shed a little tear.
NeXTSTep
You're probably less likely to see posts saying "I'm glad they kept the UI the same, I already liked it".
Perhaps it is the annoyance of change for change's sake. If you buy a car and then the interior changes after an update, it'd be pretty annoying, yet that is what is foisted on us.
> No freaking way to disable it.
Um. https://imgur.com/a/ydt3MN5
Liquid Glass is more than just the transparency. There are many other changes that cannot be disabled that way. Not to mention the overall lag compared to Sequoia (I skipped Tahoe entirely so I had a direct knife-edge transition). 27 is less bad than 26 but Liquid Glass doesn't make anything better than prior UI and makes several things worse, even if you turn down the transparency.
I imagine that the UI changes since Sequoia regarding the giant buttons and whitespace are because Apple plans to release a touchscreen laptop (most rumors say it will be announced tomorrow) so they've decided to make the UI touch friendly at the expense of bloating the UI for everyone who doesn't have a touchscreen. Microsoft made this same decision with the Windows releases after 7.
The rumors also said it would be announced today, so we'll just have to see. Even a broken clock is right twice a day, after all.
Press event is tomorrow. I don’t think a touchscreen laptop is a good idea, but I guess we’ll see whether Apple can do a good job with it.
A touchscreen laptop is only a good idea if it can also be used like a tablet (with a detached keyboard). Otherwise it's useless.
Why? My SO has a Windows laptop with touchscreen, touchpad, keyboard and mouse. She uses the mouse but sometimes pushing a button by touching the screen saves time. If you think about it, by the time you reached for the mouse and started aiming for the button you already pressed it with your finger. The touchpad reduces the travel of the hand but not the time to aim. Hitting something with a finger is always faster. She is missing that on her work laptop.
> If you think about it, by the time you reached for the mouse and started aiming for the button you already pressed it with your finger. The touchpad reduces the travel of the hand but not the time to aim. Hitting something with a finger is always faster.
Depends on how accurate you need to be. How fast/confident you can be at hitting with a finger can easily drop off a cliff.
> The touchpad reduces the travel of the hand but not the time to aim.
This is likely true for casual users, but for veterans this is a skill issue. My touchpad accuracy/speed more than matches most folks touchscreen accuracy/speed.
It may also reflect the quality of the touchpad. For a Windows laptop, it's virtually guaranteed to be of lower quality than something offered on an Apple laptop.
Reaching up to touch the screen is slower than reaching down to the trackpad. Also consider trackpoint.
Skilled users can make the reach and the motion one. You don't need to stabilize your palm on the rest to use the trackpad. Same way I don't need to use F and J for the keyboard (and kinda wish I could remove those annoying bumps)
I give you an example on this page of comments, so you can check if my informal results apply to you too.
I am moving my finger on my touchpad to move the pointer to more or less match what I am reading. I am always close to a Reply button or a vote one. There is no point reaching for the screen with my hand and touching those links. Furthermore they are way too small to be pressed with a finger. They are probably too small for a pointer too, but that's another issue.
Hoever, when I open an empty tab, type new in the URL bar and FF completes it to the URL of this site, and I press return, I usually end up with a Login link in the top right. If I reload the page (ctrl-r) I am logged in. For some reason either FF is not sending the session cookie or HN is not doing what it should on a cold session. This is also another issue (sorry) but let's say that I want to login. My pointer is far away from the login link. My fingers are not on the touchpad. Is it faster to move my right hand towards myself, touch the touchpad, slide the pointer to the link or moving that very same index finger to the link and touch it? My laptop does not have a touchscreen but when I used a tablet with a mouse and a keyboard I reached for the screen many times. Of course a mouse is less efficient than a touchpad, except for games, but I also miss the touchscreen sometimes.
And I would use if to press the reply button below this line I am typing. Let's see where the pointer ended up while I was typing... At the top right corner. Moving it to Reply and clicking is slower than touching the screen.
I used touch on Windows to test multitouch and pen input support in a familiar environment. Less hassle than grabbing a tablet, setting up remote DevTools, etc.
Yeah I think this is probably right. I use a 13" MBA as my primary machine so space is crucial and Liquid Glass wastes it.
The way things are going you will be able to run your favorite flavor of Linux on those pretty soon. It will be the best of both worlds: the best hardware and a good OS. Hopefully Apple doesn't take steps to block it.
Whats so forced on you? I'm on still on Sequoia and no one forced me to update to latest OS. Everything still works and I get security updates.
They're going to force us soon. When getting a new device I don't think you can install an older version either.
I got a second hand M1 air like two years ago for 200usd with 1tb hdd and 16gb ram. I got it after my intel Air died after about..12 years? (Edit: the ssd died, which is removable… but literally had a plane to catch). All my pros have died within a few years. Air all the way. Never looking back.
I wonder if this is from not trying to get maximum performance, and instead focusing on efficiency? But I've noticed this too. I definitely prefer the Air form factor over larger models as the primary benefit, but the durability seems to be a big benefit too. Fewer ports, thinner shell that seals tighter and can't flex as much due to the curved metal instead of flat surfaces... everything about it just seems like a better design.
Yeah the formfactor, the noise, the heat… even the performance seems better except for what, benchmarks..?
No fans though. I like being able to crank them full blast. I'm not sure this pro would ever actually run the fans all out with my usage with the existing fan curve. Apple always favors it running quiet and hot, but I don't mind loud and prefer cool palms. I grew up with far more noisy rigs.
> All of my pros have died within a few years
What are you doing to them?
I still have MBPs from 2015 and 2018 that have not 'died'.
Just use. There were two models that had issues with the gpu desoldering, 2012 and 2015 iirc? I was lucky enough to have them. I put the mobos in the oven (180deg for seven minutes?) and that would fix it. For about a month… and then 2 weeks.. and then… I actually have a 2018 that I borrowed from a friend after my 2013 Air ssd gave up, it is still working I think, but its twice as slow, at least as my Air was. Its so weird. The Air stayed ‘snappy’ even after my friends Pro’s would become too sluggish. Maybe because I never update ;)
Sounds like user error tbh.
Its from when they switched to lead free solder. Its a known issue.
Did they really wait that long? I thought most commercial products went lead-free in like the '80s or '90s. I have an old Sony PVM with known issues due to low quality lead-free solder, fixed by reflowing/adding fresh solder. From what I can gather it was released in 1994. In the same vein I'd have expected lead-free to be a bit better by 2012.
Mid 00's, its part of why Xbox 360s were dying left and right.
Similar experience. In active use in my house:
The longevity is incredible.
My first gen 5k iMac (2014) is our kitchen machine these days. Still a joy to use. The final gen (2020) model is still my main machine.
I almost wish one of them would konk out so I can get an M series, but then I wouldn’t be on a 5k iMac anymore, which would suck. Love these machines.
At least two of these are now on completely unsupported OSes that are no longer receiving security updates (and lots of latest versions of apps won't work). Especially given the uptick in major CVEs of late I'm very nervous about reconnecting any of these machines to the internet :/
no one is going to hack your laptop that sits behind your router at home checking emails and browsing the web
I get that this is hn and we're paranoid about everything, but come on, what CVE is a threat to joe sixpack here where he should drop $2K on a new machine to protect against it?
Opencore legacy allows you to install newer versions of MacOs on old Macs. Or Linux runs just fine on the older Intel Macs.
That's what Linux is for I've got a 2011 MacBook Pro running Debian and its still perfectly usable(after maxing it out to 16 GB RAM, installing an SSD, tossing the CD drive, repasting the CPU, and replacing the battery). I'm curious to see how long it can keep on trucking, at some point the charger may be the difficult part to procure; it’s not very robust.
It's a great point! The 2015 has been a linux machine for ~5 years, and the 2012 is my wife's laptop, used once every couple of weeks. Trying to get her to upgrade to a MacBook Neo for this exact reason.
In 2024 I sold a 2010 MBP that still worked. I beat the hell out of it in college dual booting and cooking the cpu playing Trackmania for 12 hours at a time or running Solidworks, then after I graduated it sat closed behind my stove sharing its wifi connection (my neighbors unprotected wifi) out the ethernet port and into my own managed router for about 4 years. Up until the last days it was running Windows as a host for controlling my old not-network-connected 3d printer. I definitely got my money out of that machine, someone paid $150 for it.
They fare differently in dusty environments, for one. Airs being sealed do well.
The last few models of Intel Mac Pro that I used were mostly good at getting really hot.
Dropbox would fire up to do massive walks of the filesystem tree and the fans would take off. I think the thermal interface material on the GPU got cooked at some point. It would be a relatively easy fix, for me a couple of hours, but I chose to fix it with an M1 MacBook Air.
The old airs were legendary. My 2012 still runs just fine despite being beaten to hell. Would still be a great dev machine for non-GPU workloads if it weren't for a lack of OS updates. Have been very happy with my M1 Pro as well though.
the 2012 was a really solid generation (the 2010 as well. i still have both, not running osx anymore of course)
Another 2012 Air user here.
The battery doesn't hold a charge anymore, but as long as it's plugged in, it keeps chugging along.
Wow turns out there's a whole community of us! Same boat. I usually recycle, hand me down, trade-in but have really had a hard time letting that little one go.
I also have a 2010 air still in service. Replaced the battery once with an Amazon special. It’s due for another but I just keep it plugged in mostly
I replaced my original in a 2012 pro. The replacement died so I run it without a battery.
> The old airs were legendary.
You gotta be kidding me. The old M1 Air had the "bendgate" where several people, me included, saw their screen die overnight for no reason. Tens and tens of pages of people complaining about it only to be with "you're holding it wrong" answers. Insanity.
I think Apple fixed that mess with the gen after that one.
We're not talking about M-series Airs.
incredible deal
Yes. The screen broke after one week, but the seller even replaced it for free. They seem to go for double that today now :/
I'd still buy MacBook Airs even if they cost more than the Pros. The lighter weight and the silence of having no fan are worth more to me than the extra performance to me.
I think differences were greater pre-M1, these days Airs are like low budget config Pros missing the fan. I highly doubt they'd last any longer. Maybe you could argue that with the best CPU/GPU options used in the MacBook Pro it will degrade faster than any Air, but if you spend about the same on each, I expect the Pro to win.
My son's M1 Air CPU died suddenly. Maybe it was a thermal death from overheating. Not sure. But all the data on it was lost because the storage is soldered on. PSA to make sure devices with non-removable storage are backed up!
(We replaced it with a Framework... with removable storage.)
Isn't icloud backup on by default??
Kind of, but without a sub you only get a trivial amount of storage.
Sure, if there's anything worth subscribing to, it's that first $1 a month for your critical files.
I'm not aware of any control over what gets backed up.
I'm assuming that all of my local work would be lost. But I'd still have the sixteen million duplicate photos.
iCloud's priorities are not a great fit for preserving my source code repos. Fortunately those aren't very big.
But when my son's Mac dies, his huge Logic Pro projects are toast. And he is a few thousand miles away now.
In Windows, by default, Microsoft tools put projects in ~/source/repos, which OneDrive, in turn, by default won't touch so they'll be lost if the SSD fail and they're not in git.
However, OneDrive doesn't seem to even know how to handle source files (it doesn't even attempt to honor the gitignore), so rebuilding a project in folders covered by OneDrive results in flurries of temporary files being uploaded/updated to the cloud, hogging the bandwidth, with a bonus of locking files, crashing builds, or even reverting changes silently.
One would think backup tools, even if primarily for normal users are still literally made by programmers, at least consider the use case of programmers instead of just ignoring them entirely.
OneDrive, Dropbox etc are not backup tools, they're file sync tools. This really should be common knowledge.
i'm so tired of seeing this discouraging crap - "that's not a backup, LOSER"
file sync is backup against the things most people have to deal with (hardware breaking, theft, house fire, residential lightning strike)
>file sync is backup against the things most people have to deal with
I would say accidentally deleting a file is more common than a house fire, so NO, file-sync is not a backup nor is RAID (block-sync) or clustering (data/state-sync).
A backup is/should be a non-modifiable snapshot of data/files maybe even a state.
I give you an example:
You have a password-safe (as a file). Your computer blows up, but you are safe, because your password-safe is synced on Dropbox....but oh my god what was the password to Dropbox??? Your restore is gone because you forgot to "backup" your password-safe.
However you are not a Loser, being a Admin is a ~lost art and a really good paid one (again).... ;)
> I would say accidentally deleting a file
It goes to your trash can/recycle bin for a while. That's literally what it's for: to guard against accidental deletes. Very few people accidentally delete files from their computer's drive anymore. The only time people do this in 2026 is accidentally formatting an SD card that had their camera's pictures on it or something.
I guess I'm not tracking. I put all my source folders in my Documents folder on the Mac, and check the checkbox to have that sync to iCloud. It works quite well.
I also do a bunch of video work with Da Vinci Resolve, and those project files are in the same folder. I don't keep the clips there, those are on an external disc, but that's just because I haven't gone from the two terabyte iCloud plan to the six terabyte plan.
There is no iCloud Backup for Macs, only iOS devices.
There is iCloud Drive sync, but that only syncs your Documents and Desktop folders, but any other files are not synced.
iCloud will sync your contacts, Notes, Keychains, and other data on macOS.
Surprising somehow, to me. One would imagine Apple could make a good amount of money by selling storage and thoughts-off safety for normies? And catch up to what Google offered like then years ago already with ChromeOS.
Yeah I've always thought it was weird, especially after they cancelled the Time Capsule, leaving no official Apple product to back up your Mac. I figured it was because the max iCloud plan was 2 TB and Apple sold Macs bigger than that, but then they increased the iCloud plans and still didn't do it...
> PSA to make sure devices with non-removable storage are backed up!
Really this should be devices full stop. Also, "backed up" should include no fewer than two backups on different media and in different physical locations e.g. Time Machine to an external SSD _plus_ continuous cloud backup.
And restore tested.
The more you people keep harping on this ultra robust backup strategy, the fewer everyday average people are actually even going to attempt a backup.
1/ You're on HN, the audience isn't "everyday average people"
2/ This is like complaining that climate activists were harping too much 30 years ago, and that's why nobody listened
It's getting more average every day. QED.
> 2/ This is like complaining that climate activists were harping too much 30 years ago, and that's why nobody listened
this is a serious issue though.
if you harp too much about something, it becomes background noise. people get desensitized.
if you harp about too many things, it fatigues people as well (ever heard of "alert fatigue"?).
so yeah, harping too much, too loudly about too many things is a serious and the society should really start looking into this
(edit: i am aware of the irony of harping about harping)
What would have been the right amount of harping about the environment though? Or is it always going to be a pointless exercise trying to convince/alert people?
Or is the only way to do that successfully, doing it subconsciously/covertly like the state actors are swaying public opinion? :-/
Unfortunately I think the sweet spot is to say it once or twice, go away, and then come back later with "I told you so"
No pain no gain also goes the other way, many times people don't learn until they get slapped in the face.
> "I told you so"
That'll be fun when half the world is migrating due to extreme climate events.
1. And yet we have someone on HN reminding everyone to even have a backup in the first place so let's not assume everyone on HN can't also make "average people mistakes".
2. You mean the climate activists who used to risk their lives back in the 80s on open oceans instead of gluing themselves to streets? Yeah, the times have changed.
I agree, if it has too much friction it's more likely you will postpone it. Any option, either just a cloud backup tool or a spare HDD or SSD with CCC or TimeMachine is better than nothing. And you have to plan for the most likely case, e.g. spilling beer on the laptop vs the house burning down.
For the house burning scenario, a cheap strategy is having one or two old HDD's in USB cases, located somewhere away from your house, in a place you regularly visit and trust security-wise. Then you can just make a habit of dumping the files there whenever visiting. Most important data of mine is safe even if internet goes down and somebody nukes the city I live in, for zero recurring cost.
While I was expressing an ideal setup, it isn’t all or nothing. A single backup is still better than none at all.
I suppose I didn’t specify exactly what I would recommend to someone who thinks it’s too hard, so here is that:
1. Install Backblaze. This is basically fire and forget. 2. (1) is enough for most people, but if interested (and on a Mac, which is what I’m familiar with), plug in an external SSD once a week or so for Time Machine. Setup involves clicking on the drive in the setup wizard and specifying an encryption password.
I don’t know about “ultra” robust, but this gives you protection from accidental deletions, natural disasters happening to your computer, hardware failures, and corruption from software bugs. The ongoing maintenance is plugging in a hard drive once a week.
If that is too much effort, just (1) provides natural disaster and hardware failure protection. It helps somewhat with the others but Backblaze only saves versions for 60 days and that’s a tight window to notice subtle corruption of e.g. your photos.
if you want to talk about 'average people' go hang out on reddit. How many times have 'average people' been brought up here explicitly to shut down a discussion. It is reductive and insulting
how many good things have 'average people' (and the avaricious wanterpreneurs chasing their dollars) ruined? Nearly everything!
This is a not-very-large subset of the ways you can permanently lose data which exists in only one copy. Have a Time Machine. No step two.
> We replaced it with a Framework... with removable storage.
Cool. How are you backing it up?
Can't a decent shop desolder it tho?
Key is held on a separate chip unfortunately
That's a big ask. Probably only work the few hundred it'd take to do that, then get the data off it, if you had some super important documents // crypto wallet.
Is it really? In China you can upgrade your iPhone storage this way. Sure it's not for your average phone case store, but it's not exactly rocket science either.
Unfortunately not really a lot of places to do this in the US. And unlike shipping new devices from China to US, shipping used devices from US to China and back is prohibitively expensive.
All the data is encrypted and the key is held inside the secure enclave within the CPU. If the CPU fails then the data is unrecoverable.
encrypted by both a boot password and a secure enclave held key
And the owners of the computer can't get the key in the secure enclave? Are they part of the threat model?
“Some person with physical access to the computer” absolutely is part of the threat model.
Some person, yes. The owner, no.
The standard way you distinguish "adversary with physical possession" from "owner" is "do they have the owner's passphrase".
I am not familiar with Macs so I might have been wrong from the very start, but if the owner put the key in the secure enclave then the same user should have no problem at putting the same key in the secure enclave of another Mac. If the user can't write into the secure enclave, then the Mac is protecting the data against the very same user that is supposed to own those data. It is protecting against wrench attacks too but what's the most frequent scenario?
Are you suggesting that both keys should be required for decryption?
How would you restore the backup if your secure hardware device is lost/stolen/damaged?
The backup would live someplace other than on the broken machine. It would be encrypted some other way (or maybe not at all, depending on threat profile).
No point in encrypting those family photos burned to disc.
you should be backing up all devices with removable storage or not, SSDs can also die without warning.
Sure, but in this case your cpu dies or ram dies or ssd dies, in all cases you loose data, in an upgradable system you loose data when ssd dies. Different risk factors.
that's just a false sense of security. A machine should be treated as a whole. Separate components only matter economically, not for data security.
My macbook updates my documents etc to my NAS, and the NAS syncs to crucial folders to the cloud (not everything for cost reasons)
You should always assume your device can be stolen, you can spill fluids on it (or a guest), a cat can pee on it, an electric shock may kill it... No matter how great care you take of it, shit happens, data needs to be safe.
The SSD dying standalone vs the system dying doesn't change at all for data security. It's not a different risk factor
It's the same "risk" but a completely different probability for anyone technical because any minor failure is now a critical failure.
I wouldn't choose to buy a car I use every day that's 100x less reliable than a normal car.
Hard drives failing is a far more common problem in my experience. As is theft and loss - one should always treat a laptop as if it will fail at any moment and keep good backups.
If your looking for arguments against soldered storage the facts ssds are wear items and will eventually need replacement is a much better one.
If almost ANY failure means the laptop is basically bricked then even with good backups it means days or weeks before being back up to speed, having a working laptop with recovered data, esp. when working on highly confidential data as agreements can mean you cannot upload the data to a cloud.
If given a choice, I prefer a system where a HDD/RAM/GPU/CPU/PSU/Screen/Port/Battery/Keyboard failure is barely if at all disruptive to one where all of these effectively brick the system.
SSDs are a more common failure point, as is theft or loss or damage.
The point is it shouldn’t matter if the cpu or ssd dies because you have backed up everything already and are prepared for anything to fail or happen to the device rather then “oh the ssd is removable so in these select few situations I can recover my data”
How removable storage will protect you from burned down laptop? Do you plan to keep SSD in your pocket while you are sleeping?
The removable storage will presumably not be encrypted the way the soldered storage is
what does that even mean? why would encryption depend on how storage is attached (soldered or plugged in)?
It's unrelated to the physical connection, but the drive encryption used on the built-in drive keeps its keys in the CPU's secure enclave, so dead CPU == data lost, regardless of whether the ssd itself is damaged.
Obviously if you used the same encryption scheme on a removable drive you'd hit the same issue - but most people don't do that, because storing the encryption keys in the Secure Enclave makes your drive effectively no longer removable.
Even if the storage is soldered on the data can still be saved if it's not the storage that died and even then, it could be possible to retireve at least part of the data. Get a data recovery specialist.
That’s pretty expensive though.
I bought my M1 Pro the week it came out and it was unlike any computer I ever had, expectations, compute, battery life especially. It was just a joy to use and not worry about chargers. I'm so glad they went this path.
Donate $3k… to some charity that’s going to use it to justify their COO’s $200k/year salary. GTFOH!
you can probably get the battery replaced for a better speedup
My worry is mostly storage. SSD do not last forever
There were so many dummies out there saying those early M1 macbooks were doomed to fail early coz of high SSD swapping
Apple with their SoC really changed a lot of people's minds about hardware.
I had an M1 Max machine with 32GB of memory and upgraded it to an M4 Max machine with 128GB of memory for local AI model experimentation.
It's sad that M5 (and M6) are so much faster at prefilling and other LLM-related operations, but I'm happy with the computer I got and will keep it at least 3-4 generations.
They really have radically improved the situation for prefill speed! But the generation speed is behind enough that you'd notice. An M4 Max is a great choice I think.
Stay strong! I went from an M1 to M5 generation, and it was a nice bump, but I could have gone another year without any significant impact on my quality of life/work.
second this anecdote
Same. Only because I broke my M1. Performance wise the M5 barely feels any faster to me in 90% of tasks
still rocking the m1 studio, shows no signs of stopping
Same here! M1 Max going strong. I'm curious to see what they put out tomorrow but spending over 5K in a machine is a pretty rough pill to swallow
Right there with you
I have tried an M4 Macbook a while ago but my M1 Max Macbook still feels plenty fast. Hope to keep that for a long time.
I have a M1 pro for personal use and an M4 pro from work and I don't notice any difference switching between them. The M1 is still more than enough for most computer usage.
The only reason I upgraded from an M1 Pro to an M4 Pro is because I wanted more RAM. The processor itself was still plenty fast.
Yeah. M1 Max with 64gb remains more pleasurable than my work M5 with 48gb.
Im on that same cycle. M1 pro max still running strong though. 3 cycles might be hard though...
I just bought a m1 Air with 8GB as secondary device to the m4 mbp that my employer provides since they dialed up the spyware lately. I should have bought a m4 Air last November but I hope the m1 holdes until the rampocalypse is over.
I have an M1 Pro as well but my next computer will be a split setup mainly because I realized that I'm most often working / focusing on hard things while at the desk and just browsing / checking in on things when I'm away from the desk so I'm thinking something like:
- a low end Macbook Pro or Air
- Mac Studio, mid-to-high specs
Why not just continue to use your m1 pro as your low end client? Save a couple hundred.
that's the plan for now, I just haven't had the M1 Pro slow down/hang/lag enough to give it up. It's a mid-range M1 Pro with 32GB of ram, works for me for dev work, I don't run any local LLM or resource intensive things on it. Battery only has 250 cycles on it so it's not like it dies off-charger.
I have M1 Maxes (one for work, one for home) that perform as good as new, as far as I can tell. Only thing they lack is an Nvidia GPU. I might pick up the new uber-studio though tmw when it launches (presumably).
M1 Pro here as well. I will buy M6 Pro Mac Studio o Mac mini.
I've been using my M1 Max 64 GB for 5 years now and it's pretty zippy once you rewrite some slow software. I have an M4 Max with 128 GB but I never ended up migrating to it. That's how good it is.
I do enjoy the fact that if software annoys you, it’s easy as hell to simply generate a better version.
Which software have you rewritten for speed?
Text viewer, password manager, console/browser hybrid, specialized text editors, household agent, task manager, kb, that sort of stuff.
Not op here, but I rewrote the file manager giving claude only two goals: cold startup below 100ms and 30mb nikon raw preview also below 100ms. One night in a loop and it hit around 20, 30ms respectively. My macs now feel uber fast.
only the M1 Max can't do nested virtualization (needs M3+)
you should, I was surprised how seamless migrating was. I just connected both devices with a usb c cable, waited an hour, and everything was moved. Including application settings etc... It was completely painless. Just needed to login to some new websites etc.., but it took me less than 30 minutes of my effort to migrate, mindblow (total time was 2 hours because it copies all your data, but you don't have to stay with it)
My M2 pro is perfect still other than the fact that the 16gb of RAM is killing me.
I have an M1 and an M5 MBA. For the most part, the M1 is doing extremely well, but imo this is the year where M1 is showing its age.
I think I'm on the same computer, the M1 MBP with the touchbar right? I was eyeing an upgrade for a while but now that I've transitioned to running agents on a home server instead of locally, I think I'll be putting my money into a ram upgrade for that box instead. My macbook is basically a dumb terminal at this point, the only reason for an upgrade would be battery life. I might get an air or even a neo if this thing dies.
They will pry my M1 Pro Macbook from my cold, dead fingers.
My M1 MacBook Pro has been the best hardware investment I’ve ever made. I had planned to update it in 2025 at the four year mark, but it’s still going strong so I don’t see a reason. The new plan is to go to the Apple store the very day that it dies.
I also have a high spec M1 and was delaying upgrading, but eventually I needed more RAM.
My M1 is still fast and enjoyable, but I regret to inform you that the M5 actually does feel faster in day to day use. I thought for sure I wouldn’t notice, but the difference is there.
When I do large builds the difference is also bigger than I was expecting.
IMO Apple knows a ton of people are doing this, especially with the rumors on memory bandwidth aka LPDDR6X bringing huge generational uplifts[0], which is why it seems they haven't brought down trade-in values from a few weeks ago, at least for my base spec M2 Pro MBP - Still $835. Pretty compelling to do that and pay just over $100/mo for a year to get a M6, given I've already gotten >3 years out of it.
0: https://news.ycombinator.com/item?id=48681392
Doesn't the high resale value of Macs not make it worthwhile to hang on to your old one as long as possible?
Also why would you have to spend $6000? You can get a refurbished 16-inch M5 Pro with 48 Gb RAM and 1 Tb SSD from Apple for $3000.
That's funny, I bought the config you are describing, but with nanotectured 14" instead of 16 new in July from microcenter for 2700 after tax.
Why wouldn't it make it?
Mine is running great and it has 24/7 use.
M1 Max 64gb here, running Qwen 3.6 35b (unsloth 4bit) and it's plenty usable, and cheap used. Best bang for buck.
heyyy same, checkout kapitsa as the inference engine, gets 145+ tokens/sec
https://github.com/nace-ai/kapitsa
this machine just keeps getting better and better in the inference world. the memory bandwidth is too good, newer macbook pros aren't that much faster in memory bandwidth, many are slower. not enough to upgrade yet as they're mostly horizontal upgrades, the improvements in other areas are adding up like wifi bandwidth and protocol, NVMe speeds, webcam
if they do the M7 correctly: which is just everything their current models do plus way faster memory bandwidth and much greater memory possible 192GB+, then it will be a really great and timely upgrade
Here are the numbers that got me buying a used M1 Max for LLM work.
Only Max are worthwhile for LLMs. An Ultra is just two Mac chips on a die, but I think they're just used in Mac Studio.
I'll check it that inference engine! Thanks!
Edit: it doesn't give numbers for unsloth/Qwen-3.6-35B-A3B 4bit, but I'm using mlx-serve right now and getting petty good numbers (and it took a while to tweak everything properly!)
it has just a few models it can download, it's really qwen and apple silicon pilled
it has unsloth on the list of 3.6 35b variants only on the 2-bit unsloth UD-IQ2_XXS GGUF variant
I'd say let it down 4-bit Qwen 3.6 35b again and see what how it performs in both benchmarks and your preferred conversational and agentic harness
it's Kflash verifier is competing against MTP and Dflash2, thats where the additional novelty is
> improvements in other areas are adding up like … webcam
Has the webcam changed or improved in the last couple years? It seems to still have the same webcam specs as the M4?
no it’s stuck there, M1 has a 720p cam with a smaller sensor thats noticeably bad in low light, and the refresh has occurred twice to 1080p with larger sensor and 12MP but cropped to 1080p for audience tracking
That’s what even low spec computers used to cost, not even adjusted for inflation. I don’t think spending $200-400/month on a hardware budget that enables me to earn 50-100x that is that onerous, we just think “six grand for a laptop???!?” But it’s ridiculously cheap for what it is. We wouldn’t be similarly shocked at a $6k car, for example.
I figured my direct family could use an extra Macbook, so I decided to "peel" my M1 Pro away as a gift to my mom. It stays close, I can use her unused space as backup, provide ad-hoc Tailscale to her (way smoother than with her crufty Windows machine), and I get a nice upgrade to my current workflows.
Having said that, I did plan on sticking it out 7 years. Ended up bailing on an M5 Pro!
Also M1 Pro here. Any reason why you're planning to upgrade at M7 Pro? I still don't feel the need to upgrade, as much as I wish I could justify it.
I like to factory reset my devices every couple of years, it really gives them a new lease on life. My M1 Pro feels as smooth and fast as the day I got it.
I think the rumors are the M6 and then later M7 Pro/Max will be put into a new chassis with a touchscreen and no notch.
M1 Max, 32GB, 1TB; still not feeling the need to upgrade, especially when the upgrade is as expensive as they are now.
same
Why would you need a $6000 computer for dev work? For toying around with local models I would understand.
Standard issue at our company. There's just some apps that are huge, especially legacy ones.
For the ability to run our full server-side application stack locally, including all services and as much of the test suite as possible, as well as the profilers, debuggers and telemetry tools to debug issues. Way faster to develop when you can have a change running live in about 3 seconds instead of 5 minutes, run targeted tests in seconds and a full test suite in a few minutes.
That's fair, I've never worked in a domain where I need to run that locally. Is there a reason to not just SSH into a workstation that can run it?
Not op here, but: i have a build box i ssh to, but when on the go: planes, taxis, etc. its annoying to constantly have to connect. Slightly. Having a fallback, fast local laptop is a convenience. Tho I do travel a lot - usually >100 flights per year.
When I was in the market for a new machine, I had the same attitude. As I was about to hit "Buy" on a standard issue $2k MBP, my wife said, "What if we got you a top of the line machine?"
I replied that it would make no sense unless I was going to start running local inference or training my own models locally...
2 years later, and this MacBook Pro M4 Max w/48GB is the best $4k I've ever spent!
Big test suites. I have one project hitting 35 mins on m1 max, vs around 12 min on the m5 max, vs. 7 min on the linux box I keep in the basement. Unit tests, integration tests, headless browser with webgpu stuff.
And, here I am with my 2012 macbook pro, still thinking if its worth upgrading.
Are you on the model with the CD drive or the high resolution "thin body" version? I graded the ram in mine and put in an SSD, and it still works beautifully, but I'm now on now on Linux with it, and have been for some time.
My M1 screen just cracked :( Debating to get it fixed (it’s a 32GB 2 TB machine). Crazy that I likely will spend 500€ to get a super old machine repaired just so au do not have to spend >3000€ for a similarly equipped new one.
Happily running my M1 Max, but have recently been annoyed at my decision to go with the off-the-shelf 32GB version at the time. I recalled telling friends at the time that the ridiculous memory bandwidth should offset that, so I ran some tests on memory pressure while running all my mess and, well, whaddya know, everything is actually tickety-boo. No memory pressure for the most part. And this is with big rust projects, Orbstack builds and runs.
The engineering that went into the M1 was amazing, I truly think that now, an M1 Max was the biggest bargain ever in my entire computing life. £3,299.00. In contrast, the 64GB i9 I bought only two years before was £4,149.00.
The “on-device AI” with the standard M6 seems to be slightly faster than the M1 Max of five years ago. All the rest seems to more or less match. Hope the M7 Max is yet another leap that serves many users for years to come.
The M1 Max MacBook with 64 gigabytes is still a great machine and fairly affordable on the second-hand market.
(Returning an i9 to wait a few more years on the M1, any M1, was absolutely worth it)
Got an M1 Max 64gb for $500 last year from a tech company inventory sell off. Had just 2 battery cycles, basically brand new. Been loving it.
Still with macmini m1 8gb here. still working OK taking care of space and RAM.
Haha same here. The M1 Pro is such a powerful machine and it still works great. I did buy a MacBook Air M5 for non work, and I think it’s about as fast as the M1 Pro, but the latter has 64GB of RAM so I can actually work. Will probably request a new one next year though, it’s time.
I’m hopeful mine survives until RAM and SSD prices return to Earth. Unless I drop it I suspect it will. It’s still going strong. Maybe the M8. Hopefully chips and local LLMs have developed to the degree that local inference will be good and fast.
I saw a decent listing for a 16" 16gb m1 pro macbook the other day, and i was seriously considering spending those money.
I have a 16" m4 max from work (36gb) and is an incredible machine (build quality, incredible display, overall feel). I also have a private macbook neo which should be similar to M1 in terms of performance but it's limited to 8gb of memory.
The neo/M1 performances are sufficient for my needs (for private computing) and chances are the M1 pro actually performs better than my neo (if anything for the more abundant memory).
I might to back to the listing and try to haggle the price even lower, considering a new generation of macbook pros has just been announced.
I gave away my base spec (8gb ram, 256gb ssd) M1 air just a few months ago and it's still working fine, I just needed something higher than base spec for work but the new owner of the M1 loves it, for general purpose use it may as well be an M5.
I do wish they had removable storage though, I originally got that Macbook as it was the cheapest way I could try out MacOS and M1, once I decided I liked it, I was stuck with 256gb or spending all that money (and then some) again. This time, I went with 1tb.
I'm also on an M1 Pro, trying and failing to find a reason to upgrade other than wanting a "shiny new toy".
Just replaced the keycaps on my M1 for the second time to make it feel new-ish again.
Still a great machine, shame those keycaps wear so badly (developing shiny greasy-looking patches)
my intel to M1 Pro upgrade was like stepping into the future, still on it and my only motivation fro swapping computers is i dropped it from my lap when i fell asleep after a long day with the kids, and the n key hasn't worked for a year
authorized repair shop says 4 days to fix it and it's my only laptop, so i go without :(
Indeed, pricing has gone a bit out of hand. I got the top-of-the-line MBP when I was starting my PhD at the end of 2013, for something like $3,200. You couldn't get a more expensive consumer-oriented laptop at the time. It's mind-boggling that the equivalent purchase today costs around twice as much.
I think it's things like this that are a great argument against soldered ram. Especially with memory shortages.
(Obviously there are other reasons to solder like performance)
Exactly on the same boat: I don't see a big improvement until I can run a decent LLM on my machine. But I'll wait for it to reach the $2500 price tag.
Until then, then M1 Pro will be fine for my usage, which is already more advanced than most.
My ideal scenario:
- 2027 the AI bubble bursts
- 2028 new memory factories reach the production stage and fail to meet demand, so we have a glut of memory and rock-bottom prices.
Me too, running a 16 inch M1 Pro. So glad I went for 32GB as that'll add to the longevity. Providing no hardware failures I can see at least another 4-5 years. It's been running silky smooth up to now even with the Golden Gate update.
I mean, you don't have to. Do you need a Mac? I make mobile apps and Xcode is only on macOS so I definitely need it but most people don't, there are Windows and Linux options that are great for fractions of the price.
I think he wants a machine that can do local LLM
I'm always surprised that devs need such high spec machines just for developing software. Unless you're developing games or scientific computing software or similar, I din't know what the power would be used for.
My dev machine is an i5 of some description about 6 years old, that's about all I could tell you off the top of my head because I've never worried about needing slightly faster compiles. (Ignoring running local LLMs of course. Currently they can't compete with Opus 5.5)
>scientific computing software
In my experience with this, most of the work happens on a HPC cluster. Your actual machine just needs to work as a thin client.
The IDEs for Android and iOS dev are heavy plus devs tend to run a lot of VMs so extra ram is nice.
I'm developing an inference engine and need access to the latest Neural Accelerator features, and therefore need to keep up to date with the Apple releases. M1 through M5 Max 128GB. As a broke bootstrapped un-funded startup founder this is extremely tough to afford given the price increases and all.
I'm still running a 4 year old M1 Max 32GB Mac Studio as my daily and really haven't found anything that it shows sign of slowness/age with.
That said I don't do any local AI or 8K video editing type stuff.
On upgrade I could probably live on a moderate/high spec Mac mini since the only reason I went Studio originally was the resolution/number of displays it could support.
The $2000 version is 10 X better.
I'm with you. Typing this on a 2021 m1 64gb. I keep thinking it's time to get a new one and I'd really like 128gb but justifying the spend is REALLY hard. I've had m5/128gb for a few weeks at a time and it's an upgrade for sure, but not $6k+ upgrade.
I switched entirely over to Asahi Linux on my M1 Macbook Air and installed a third-party 55Wh battery, as the original battery had fallen below 80% capacity. That machine is good to go for another 5 years.
There are no features or performance characteristics of any of the subsequent M[2-6] Macbook Air or Pro models that I care about. I'm not really sold on running models on the hardware I have right in front of me. So long as I can rent an RTX PRO 4000 Blackwell at about half a buck per hour through a European provider, I'm fine using that with Qwen3.8-27B UD-Q4_K_M for the few things I consider more privacy-sensitive.
How is Asahi Linux from a reliability perspective? Last I checked there were some issues with docking and HDMI out that would've made it tough for me to use as a daily driver.
Not a Mac user. I assume HDMI might be improving soon as AMD finally got the HDMI consortium to wind their necks in and full HDMI support was recently merged.
USB-C video still doesn't work but in my experience virtually everything else you need does.
Does DisplayLink work?
100% stable for me at the level of any other linux machine. True the USB-C display doesn't (officially) work yet but I have 0 issues with the HDMI.
I've been daily driving it for months with no blockers. You can tell Linux uses the battery less efficiently than MacOS though so that's good to know.
> at the level of any other Linux machine
So, 97% stable then? (Downvote me but it's true.)
I've been running Ubuntu on an MSI Ge66 gaming laptop, Intel + Nvidia gpu for a couple of years now. I did have some teething problems with wifi and sleep but after resolving those it's been rock solid. I'd say 99.9% stable if we're throwing out random percentages here.
While not Asahi/M1, I can tell you my desktop has been Linux Mint for about 5 years, including through distro upgrades, etc and I have had to troubleshoot it like once.
PEBKAC
97% is a high estimate based on my linux efforts.. just yesterday some kind of update, i think kernel, took out my amd graphics.. still wouldn’t pick anything else.
Funny you say this. I had my issues with Linux during the years but last 2-3 years I have been running it with no issues whatsoever on a Dell XPS 9xxx series. I recently got an M5 Max and it had more issues than the Linux/Dell duo.
Love your bait
It wasn't meant to be bait. I'd really like it if Linux was as rock solid as Mac or Windows. But it clearly isn't and it doesn't help anyone that so many Linux fans pretend it is (look at the replies!).
That said, I do have hopes that AI will basically solve this. It's extremely good at diagnosing low level bugs, reverse engineering drivers and so on. A silver lining of AI taking our jobs at least!
> I'd really like it if Linux was as rock solid as Mac or Windows. But it clearly isn't and it doesn't help anyone that so many Linux fans pretend it is (look at the replies!).
It is perhaps also possible that others have different experiences with both Windows/Mac and Linux than you do. It's common for people to not notice what's broken, or what they have to workaround, with the software they use every day. People also have different workflows and priorities that determine whether their use of a tool is "rock solid" or not. (Something I certainly wouldn't say about Windows—or Mac!)
Funny that Linux is perceived as unstable, I have to fix a Windows laptop where Windows starts up fine 1 time out of 10, while normally it freezes (not even blue screens) on the boot loading spinner, I've done everything possible (including of course running a 2 hours RAM test, and testing it with a Linux live USB) to exclude hardware problems, so what is remains is... Windows of course.
One of my coworker that uses a Mac has an issue with a USB dongle that stopped working randomly, not even with a system update. If he uses on its laptop the external monitor remains blank, the same dongle with another monitor and another identical Macbook with the same macOS version works, the same dongle with the same PC and another brand of monitor works, the same PC with the same monitor and a dongle of a different brand does work, but a different dongle of the same model doesn't. So I guess for some reason macOS saved the ID of that dongle and the setting got corrupted. Tried to reset the hell of it and it doesn't work.
I've never had similar issues with Linux. Most of the time when something doesn't work with Linux you just look at the logs, figure out what is wrong, correct it, and it start work fine again. Configuration of the system is predictable, if you install a software and then you uninstall it the system is left in the same state as before installing it, it doesn't leave files that waste disk space, as well as when you upgrade the system (yes, every Windows or macOS update leaves old stuff behind that it's impossible to delete, effectively "leaking" disk space at the point you have to perform a clean install).
Also really happy with Omarchy on Macbook. I installed the nightly kernel that has usb-c and it works really well. Connects fine to both LG Ultrafine 5k, and Studio Display. https://x.com/AGTPinsights/status/2106564829532328331
Still some tiny issues with waking from sleep lid closed, but I'm expecting to get that fixed myself with some prompting this week. The people working on Linux on Mac are really impressive (and have access to unlimited tokens), so I expect all issues to be fixed very soon.
Being able combine agents, linux software and Apple hardware is so much fun.
It's great aside from no usb-c display, mine has an HDMI port though.
It's been my primary dev machine for 2 years
It's quite stable. Been running NixOS on my M2 Max with basically no issues (HDMI working fine as well). However the lack of adequate sleep / no support for suspend to disk led to me buying a Framework 13 Pro, and I'm now selling the M2.
> So long as I can rent an RTX PRO 4000 Blackwell at about half a buck per hour through a European provider, ..
Mind sharing the provider’s URL?
Not necessarily an endorsement, but one of the more popular ones: https://www.hetzner.com/dedicated-rootserver/matrix-gpu/
I understand that managing hw can be time consuming, but if you built a similar machine it would get paid off in ~18 months.
I think the point the OP was making was that you can rent this for a couple of hours when you need it, and then it makes sense. If you are using it say 2 hours a week (or even 5 hours a week) it's going to be much cheaper than building your own hardware for a long long time.
It's a little confusing from the page, but you can't actually use those hourly in terms of spinning them up on-demand for a few hours, because you pay the € 209 - 1499 setup fee when you provision one. They're just showing the monthly price broken down into an hourly price.
They're great servers, though, I have a whole fleet of them.
How is using an external provider any better for privacy sensitive issues? European provider does not automatically mean they aren’t logging everything or that they’d nor share that with a bidder?
Out of curiosity, how do you rent such a device and who do you do it with? I'd be interested.
No the OP, but I keep a $10 rolling budget on vast.ai. I tend to use it more for training things, but equally as useful. I pay about $0.15/hour for an A4000.
OTOH if you are really concerned about privacy it's probably not the best choice (it's random people's machines).
Intriguing, thank you.
A note about privacy. AI model can do tool calls to leak your data. See https://cohere.com/blog/concerning-behaviour-popular-open-we...
I want to propose a called AI Index, if you can do say 1000t/s on processing and 50t/s on generation on a min 25B model you can call your device AI9 and you have AI7 and AI5 and AI3 with 3 being 100t/s processing and 10t/s generation on a min 10B model. Values above are on a 100k context size minimum. OFFLINE.
So any effing claim of being an AI device doesn't end on the demo of best buy laptop where you remove a dog from a photo....
Or that you have "better noise removal" or "you look better on the cam" because that's called noise cancelling and instagram filter and it's being done for ages...
Can't you just look at memory bandwidth for generation?
You also need to look at memory size (to store the model). And processing can still be the bottleneck.
It’s more nuanced than that, but yes that’s the biggest factor.
Doesn't this depend on the model and quant as well vs. the hardware and its capability? Plus memory bandwidth, amount, and layout.
My Strix Halo now can run some models through the NPU and its roughly as fast as its GPU+CPU (on some of the models - some its slower)
You can run a model at 1000t/s on a low and android phone. The output will be dogshit bad, but it'll pump tokens like nobody's business. The problem with what you're proposing is that there's no baseline definition of what those tokens are (for what purpose? of what quality?). And pinning that to a certain model or level of correctness means your benchmark becomes almost immediately obsolete.
Increasingly all I need is a laptop with a big/nice enough screen to ssh into a beefy Linux box to run agents on Herdr.
I think my 48GB 16 inch MBP will last me 15+ years at this point.
That's exactly what I am doing. I have a Dell R740 with 96 cores and 512 GB of RAM, and RAID 10 with 16 SSDs. All of my machines connect to it via Tailscale.
and I thought I was doing something special with my little GMKTec 32GB machine I use for a dev server haha
There's (at least) 2 of us! Well, mine is a little minisforum unit, but the idea is all the same.
I have a similar setup! I use a Dell R820 though, with 1.5TB of RAM and NVMe ;)
Did the same. Took a week or so to get used to the workflow after doing dev locally my entire career, but now I'm certain it's the best way to work. Even better is using a few scripts to simplify things like opening in VS Code (instead of `code folder-name` I run `codeon repo-name` from my local machine).
I did actually get a new M5 Pro/48GB machine last month, but that's only because I bought my work machine to keep my personal and professional lives properly separated; otherwise I wasn't seeing any issue running that M3 Max/36GB for a long time further.
I've just transitioned to the same setup a week ago - that's why I'm considering Air as the next upgrade, given the better portability.
One thing that bugs me mentally is leaving that machine pending on my input ("wasted") overnight etc... I'd like to arrive at a workflow where I can have agents running toward the overall goal autonomously overnight as much as they can, and I can review and iterate and "engineer" parts that need it afterwards...
For now, I still guide implementation very closely, piece by piece. Has anyone had any good experiences with workflows / skills / similar for that?
I haven't had good outcomes from letting the agent decide what to do, but if you spend time deciding what tasks to add to your backlog, then how to accomplish those tasks, when you're awake, you can end up with five or six or more projects to let your agent do overnight. planning when you're awake, execution when you're not.
I have a Hetzner box that costs about 10$ a month - I use that as my always-on box. I use Pi as my harness and they run in herdr. I can then from my Mac (bottom of the line macbook air - next laptop will be a Neo) ssh to the box and control my agents. I can also ssh into the box from my iPhone (using Termius) and continue seamlessly from there too.
Have the same setup -- got a good black friday deal last year on an AMD 395+ w/ 128gb of RAM. Still not super bullish on running AI models on it, but it makes a great dev server anyway.
Actually had claude spin up a kanban board to manage my tasks in a 'TODO -> Plan -> Implement -> Review -> QA -> Shipping -> Done' flow (with a worktree for each task) that works great!
If that continues to be the case, next time you can "downgrade" to the MacBook Air. Even today you can get one with 32GB RAM, so when you're in the market I'm sure more will be available. They are quite thin and light compared to the MBPs.
The screen on MB Air can not compete with the one on MB pro.
Yep, that's why this M6 Pro and its ancestors probably exists. It's for people whose work can run on MacBook Air specs but with the superior screen, battery, speakers, and ports.
MBPs also have a fan, so they’re less prone to thermal throttling causing performance degradation.
Well, if the thesis is that we're running the heavy work on a remote server that's less of an issue.
True. I still think it is good. My dream is for 120Hz on a future Air model. I took it as a positive sign they brought 120Hz to the non-Pro iPhones with the models last year. I can dream, I guess. And yes, I know MBP screen has benefits other than the refresh rate.
Hah, I agree. As soon as I can figure out how to replace batteries myself, I'd also be happily doing the same thing with mine.
The main problem is Apple killing support for very old models.
Don't they have pretty much the best track record of any manufacturer in the last decade?
No. They publish a list of products that are no longer supported but you have to dig it out, they are shady about it.
https://www.macrumors.com/2026/10/14/apple-three-more-obsole...
https://uptradeit.com/blog/what-macbooks-will-stop-working
https://www.macworld.com/article/673939/this-is-how-long-mac...
Microsoft is known for supporting ancient hardware. Not Apple.
I'm not sure I understand - it's been here for years: https://support.apple.com/en-us/102772
Support for security updates appears to be years longer than Microsoft.
Are we talking about different things? I meant os/security updates. I think you mean hardware?
> all I need is a laptop with a big/nice enough screen to ssh
> 48GB 16 inch MBP
That's wicked overkill for ssh + browser. A 15" MacBook Air would work for this really nicely. This is how I use my 13" Air, and for the big screen I dock it at a desk.
I wish they could bring back the 11" Air or 12" MacBook and trim the bezels for something like this. It would be an amazingly portable computer. The 11" iPad Pro with keyboard case is almost this, but you're stuck with iOS and it's actually pretty thick and heavy compared to a MacBook.
>With the new 14-inch MacBook Pro, users can connect high-speed peripherals, transfer photos, work with multiple external displays, and more with its wide array of ports.
* looks inside *
4 kinds of ports. while being generous, and including the SD card slot and the headphone jack.
I don't know anyone who would consider 2-4 to be a "wide array".
But hey, it's got USB-C and also supports UD USB-C (upside-down)
3x Thunderbolt 4 ports
It's a shame that only the Mac Studios come with USB-A ports. I know, won't fit on a MBA, but similar size and class laptops to the MBP often come with at least one USB-A.
It's just handy when you have a USB-A device and don't want a dongle, traveling, etc.
This only serves to prolong the suffering.
It's frustrating. Can't go anywhere without a dongle. At least with a Mac mini your dongles can be stationary.
Compared to Air 2 :D ?
Yea, or the ridiculous first-gen M1 with only two USB-C and an audio jack. One of which was of course used for charging.
Dongle hell is a real place.
0 USB ports while my Raspberry Pi has 4 of those plus 2 mini-HDMI outputs
> With 16GB of unified memory standard, configurable up to 32GB, and higher memory bandwidth up to 170GB/s
Apple MacBook M5 Max has bandwidth up to 600GB/s, is 170GB/s for M6 the real limit? If so, wouldn't this slow down LLM processing?
There are rumors that Apple is not releasing Pro/Max versions of the M6 with the intention of focusing these research efforts on the M7 instead.
Even then, 170GB/s is a bump up from 153GB/s of the M5. So we see a 11% uplift already.
> So we see a 11% uplift already.
Am I crazy, or are a lot more people using the term "uplift" in place of "increase" within the last year? Google Trends (for whatever it's worth) shows a significant increase (uplift?) since mid-late 2025. [0]
[0] https://trends.google.com/explore?q=uplift&date=today%205-y&...
It's monkey-see-monkey-do language.
See also: The use of "learnings" instead of "lessons."
that one and "educator" both make my skin crawl for some reason
Language changes. People at work have started saying things are an easy lift instead of a quick win. And it's not new hires, it's the people I've worked with for years. I never got on board with quick win either.
Some people adopt stuff faster than others. I've lived in NZ for nearly 20 years and called someone bro for the first time just last week while playing a sport "should have left it for me bro". Slipped out natural as. Still having bad thoughts about it.
Shit I still use “low hanging fruit”! How out of date am I?
Here I am doing the needful still.
> Even then, 170GB/s is a bump up from 153GB/s of the M5. So we see a 11% uplift already.
My M2 Pro does 200GB/s.
wtf are we even talking about here.
I believe they're discussing the base processors as a way of comparing generations, since there is no M6 Pro/Max. (the M2 base was at 100GB/s)
For a direct comparison to your processor, the M5 Pro is at 308GB/s.
It's a base M6 chip. The Pro doubles that, and the Max doubles again. And the Ultra doubles yet again. This is how the entire M series has always worked.
So the theoretical Max would be 680GB/s.
The M6 version is also cheaper than the M5 Max by a significant amount. I wouldn't be surprised if the M6 could keep up on LLM prefill (compute bound), but the memory bandwidth means that decode is going to be a lot slower, most likely?
If you want to do anything beyond getting a small single-core speed bump the M5 Pro is a better, more capable purchase. At least until the M7 comes out.
It's also absolutely wild that you apparently cannot spec one with more than 32GB, for any amount of money? Wow.
Not that I blame Apple here, for once. Seems like they're trying to cope with the limited quantity of RAM they can procure. And with the paltry bandwidth (170GB/s) of these CPUs they're a bad choice for anything but tiny local models anyway.
That's for the base model M6. The Pro can take more, the Max even more, the Ultra the most.
They're not making those models for the M6 though; they rerouted their resources to making M7 faster at running local LLMs.
If you want more bandwidth, get an M5 Pro, Max, or Ultra. If you want a lower power base model, get the M6. If you want to have your cake and eat it too, wait for the M7, reportedly in the spring.
Umm yes, the M6 chip only has a certain amount of memory controllers and channels, which means you can only have so many LPDDR5X modules.
Yes, high-capacity LPDDR5X are made and do exist, but for all practical intents and purposes, you can't buy enough of them to deliver to customers at Apple scale.
Why are we comparing the M6 to the M5 Max?
What was the speed of the base M5?
As pointed out in a sibling comment, 153GB/s.
Why are we comparing the M6 to the M5 Max?
Because the question is "Which is better for local inference out of the M6 or the M5 Max?"
If an M6 Pro/Max was an option that would change the question.
I thought we were supposed to get a design overhaul this gen?
Also framing AI use for the base M models marketing makes no sense to me. You are way better off getting a used M1 max with 64gb and way higher mem bandwidth when those are the real bottlenecks.
Apparently another announcement coming tomorrow, with the OLED panels. But according to the fpt channel on Youtube, it doesn't look any different. Just has an OLED panel with touchscreen, with massive cooling changes inside.
MacRumors is still claiming we're getting a new chassis
> Apple on Wednesday will announce 14-inch and 16-inch MacBook Pro models featuring touch-enabled OLED displays with thinner bezels, a Dynamic Island instead of a notch, M5 Pro and M5 Max chips, and a lighter and slightly thinner design
https://www.macrumors.com/2026/10/27/macbook-pro-with-touchs...
Interesting. Will be way out of my budget, but keen to see how the OLED display performs long term. Nearly everyone I've seen that has used an OLED on laptop say they absolutely love it. But I still hate the idea of touchscreen, especially with the changes to MacOS to cater for this
IMO changes to a touchscreen would've been perfect for professional artists, but only if it enabled using an Apple Pencil. Which is pretty unlikely if they don't also solve the ergonomics of doing that, e.g. with a 360" hinge.
> Just has an OLED panel with touchscreen
I must be out of the loop, you skimmed over this. A Mac with a touchscreen??
Look at the first image with the people in the field. That looks to me like a very poorly generated image, not a real image of people, and not even a high-quality AI generated image.
Is this what we can expect of the new apple? I'm not against AI generated imagery, but I expect them to be of equal quality to a real photo.
Or am I wrong, and it isn't AI generated?
The things that make me think it is just poorly made are 1) the angles of the laptop on the black case, 2) the lack of any lighting on the back panel, 3) the size of the hand on the woman in the back, and the squareness of what she is holding (if it is even supposed to be another mac).
That was my first thought, is that guy's face AI? The angle of the laptop looks off. Then I decided I don't care and moved on, but yeah, you're not the only one who thought it looked off.
Almost certainly not AI gen, it's shot with a 35mm lens (quality looks like the 1.4 L to me), probably around f8 given how the DOF falloff happens, heavy forward flood (flash or otherwise, but whatever it was it's defused) and then what's I think giving you the sense it's AI is it looks like they've over used a double S curve to bump the contrast and sharpness. https://www.youtube.com/watch?v=XZh_9HnO_4w&t=640s
It is interesting how people are confusing depth of field with AI generated. It's a shame, but perhaps it will end the bokeh conversations and disparaging of M43 sensors.
It's just a very bad picture, but every detail is perfectly coherent, it's not AI generated.
Still a baffling choice.
I’m more confused by the scene itself. Why are they in a field? What are they doing?
This is a strange release. The M6 tops out at 32GB RAM and then you fall back to M5/Pro. Kinda midway between an Air and a true Pro. Only 14" no 16" with M6.
The lack of a 16" model for the M6 is consistent with Apple's current minimum of M5 Pro chip.
Hmm, still no OLED displays? I wonder what's the holdup? (I have a fairly mainstream PC laptop with OLED display, and the display is the one thing where my MBP looks pretty poor in direct comparison).
wait till tomorrow's touch screen
I just bought a framework 13 pro, going to main that into the ground, 15 years+, with maybe 1 allowed motherboard upgrade in that timeframe, but hoping to completely avoid it and just do repairs and minor upgrades.
Just now realizing there's a difference between "MacBook Pro with M# chip" and "MacBook Pro with M# Pro chip". Was curious to see how my MBP with M4 Pro compared, and I saw it has 273 gb/s memory bandwidth compared to the MBP M6 (sans Pro)'s 170gb/s. Makes sense that "Pro" in the laptop model doesn't also describe the chip, but only if you remember that the non-pro "Macbook" line stopped in 2017 (which was the last time I went shopping for Apple laptops until buying the MBP4)
https://www.apple.com/mac/compare/?modelList=MacBook-Neo-A18...
The "Pro" in MacBook Pro has slowly degraded to "it comes with a fan so that the chip can keep chugging when it gets a hard workload, unlike the MacBook Air where the chip will throttle since there's no fan".
Screen, ports too (to be fair I've used the HDMI port about 5 times, and the card reader never).
The base model chip on the Air is binned too - 8 GPU cores on the Air vs 10 on the Pro.
If anything, Pro really means pro now. It used to just be the default laptop everyone bought because the Air was too slow and had a terrible screen. The 2020 MBP was almost as thin as the MBA and had barely any more connectivity. Now it's bulky and provides more power than normal users need.
I have a 14 inch M1 MBP running fine up to this day. 16G RAM, 1T SSD. I don't know if I ever will need to replace it.
You will if you use macos and want to stay supported. On the other hand, being an M1 puts you squarely in Asahi supported territory.
I have the same M1/16GB on newest macos, that thing still flies.
My work device is an M4/32GB. The difference doesnt feel that big, especially when you code mostly with agents.
I believe you. But Apple has a history of deprecating MacBook Pro after about 7 years
I have an M1 Air 13" and M2 Air 15", and M4 Air 14" and and M4 Pro 14" with nano texture.
I love them all. The M4 Air 14" is from work, so kinda limited in what I can do. The M4 14" Pro is kinda chunky after the Airs. The M2 15" is ideal except on a plane. The M1 13" is great as a personal device on work travel. If only I could not work...
As lots of Apple Users are in here: I'm thinking about "sidegrading", my M2 Pro runs fine, but I chose 16 GB/512 GB when I bought it and I think 32 GB/1 TB would be nicer in some cases.
Is it easily possible to sidegrade/upgrade with a Timemachine backup?
Yes. Make sure both source and destination are the same macOS version. I use 27.0.1, you could upgrade to 27.2 first though.
Then do a fresh install on your new system and then use Migration Assistant. Use a USB-4 or a Thunderbolt cable to speed it up.
yes, if your plugged into the Apple Ecosystem, upgrading to a new machine is a breeze, done it many times
I'm in the exact same boat as you, exact same specs. CPU/GPU holding up perfectly for my use cases, but 512GB is tough to manage now.
It sure is, especially with casual gaming...
I checked used prices though, and I think the market is not in our favor. rebuy.de has an M2 Max with 32 GB/1 TB configuration in "very good" condition and is asking 2100 Euro for it.
M6 is "just" 600 Euro more...
I've been pretty happy with my M1-2 devices like most posters here seem to say, but I don't have the same level of core stability while doing dev work. Am I doing something wrong with my workloads?
- Orbstack, but just for one containerised service - Laravel Herd for app environments, so all native services instead of containers with bigger VM overhead - Brave browser, profile for work and personal - PHPStorm (quite a hog on RAM) - Now agents in the mix
Had an M1 Pro 8GB but I found 8 to really not be enough. M2 Air 16GB has been a much better sweet spot, albeit it feels like the battery life on it is worse than M1. I've recently started trying to use Zed more and more as my editor instead of PHPStorm and things feel mostly stable, but I still feel like a beefier machine would be good for work purposes.
Years ago there were frequent crashes (kernel panics) i could directly relate to IntelliJ (PHPStorm, etc), so it might sound dull but upgrading to the latest versions is the first thing I’d check. In this case even a manual check on their website instead of relying on the auto updater to work.
I don't think you're doing anything wrong. I struggle with similar workloads on 32GB RAM. Compute isn't the issue, once I have a JetBrains IDE open, a dev environment, some browser tabs, Claude, and a bunch of other crap Electron apps I need the machine is already swapping quite a bit.
I have found that my M5pro gets unstable when I really stress it's memory use. I think that with 8 or 16GB you are going to hit this quite a lot more than I do with 48GB.
My big issue with the M1 generation is the lack of more E-cores, which get assigned a lot of the OS background tasks. When working at the top of the RAM limit, I'd notice issues around lag to the point that I started looking at CPU stats and found those two measly e-cores completely maxed out for hours on end. Occasionally a crash, though that could be RAM related rather than CPU.
A bit of googling around seems to point that architecture as somewhat of a problem for others, and might be why Apple quickly moved to put more e-cores on the later generations.
Something to consider. Never managed to get proper distinct confirmation that this was the problem, but a (corporate) laptop upgrade did get me out of this pickle.
The text of this product's announcement really tries to sell it as a 10x AI performance upgrade for existing MacBook M1 users.
Ok, plausible number for that use case. Striking how it seems like the M-series Macs are almost too good enough for most people.
It's almost desperate prose.
(The AI image processing in Pixelmator Pro saved me hours of work on a recent project. Two button clicks vs a day of tweaking. But even that was on my M1 MacBook Air.)
Is this an odd rollout for their actual high end MacBook Pros, or is this sitting in that space where they like to have a base MacBook Pro between their Airs and the expensive MacBook Extremely Pros?
Just the standard chip without Pro/Max makes me assume this is that base one. If so, the price is way high for where that tends to sit, but welcome to buying a computer in 2026 I guess.
This is the base one. You get a better display, more, faster Thunderbolt ports, and an honest to god HDMI port, plus some fans for cooling on long running processes that you don't get with the M6 Air (which also doesn't exist).
i think also better general support for multi screen setup
Quite messy and wonky as in Linus Tech Tips testing. Require plugging displays in right order to get desired resolution / refresh rate.
> Is this an odd rollout for their actual high end MacBook Pros
It's the same rollout they've used for a couple of generations now. First the base model chips in fall, then the pro/max versions in spring
I have to say, I really dislike the thing where the base M(n+1) is always in the market alongside M(n) Pro/Max. I appreciate the underlying chips are probably quite distinct, but from a naming standpoint its confusing as hell.
+1
makes me want to wait another one I have an M1 Max and just set up playing with an M1 (non-max) and realized it can't drive 3 displays. I recall getting Max just for this, but generally, makes me think of non-Max as a non-pro after all.
And having what's presumably the latest CPU but with potentially worse features than the last generation combined with increased prices from RAM is a deal breaker so I'll guess I'll push my M1 for another year.
Whenever I do this, I enter into a perpetual one-more-year wait loop.
When I finally cave in and buy something after 5 years I’m very momentarily sad for not perfectly timing the release window, then I realize I keep hardware long enough a half year barely matters.
It can drive 3 displays using a DisplayLink dock/adapter.
I get you, but I suspect it reflects reality. The M-with-adjective chips are M(n) chips with more cores and cache, it stands to reason that it takes real wall time for TSMC to start cranking them out with acceptable defect rates.
So Apple can either toss the early production in inventory while the whole process catches up, or they can do what they do. Better for both the company and the consumer that they do the latter.
I'm guessing the Base and Pro chips come off the same wafer, and they get 5 Base chips (some defects) for every one Pro chip (very few defects). So they run the Base-model assembly line for six months and eventually they accumulate enough Pro chips to satisfy Pro-model demand.
Curious, where's 5 come from? Is there a reason for the specific guess or is it just an example?
It was a made-up number.
You mean as in binning? High defect rates for chips but they're burned to disable non-working parts and sell them as non-pro chips and accumulate the good ones ?
The Pro and Max chips have a different die than the base model. They don't bin down to the base model. Maybe Max bins down to Pro now that those are the same die, or Pro bins up to one half a Max?
Pro chips come much later than base chips and vary significantly in ways other than number of cores (which can be up to double), such as doubled memory bandwidth and peripheral support (thunderbolt 5 instead of 4).
If you look at the platform name and the kernel code you'l realize that this is not even remotely true. The base M chips are T8xxx which are (more or less) just iPad chips. The pro/max/ultra are T6xxx and are their own new thing, starting with M1 pro which was T6000
https://en.wikipedia.org/wiki/Apple_silicon#Comparison_of_M-...
As others have said, the pro is a different mask set.
But you are right, they do bin. You can get versions of the cpu with different core counts for this reason.
I’m only seeing this with the pro and max though.
I suspect it has depended a bit on whether a new process has lower output or higher defects. High defects, they'll focus on smaller baseline chips to get better yields. Low output, they'll launch pro models first because those are lower volume. Neither problem, they'll roll out updates pretty quickly across multiple product lines.
And that should change going forward with smaller chiplets being integrated in package, vs huge Ultra chips.
And I can never remember whether Pro is better than Max or vice-versa. It’s an apples/oranges naming scheme.
Regular < Pro < Max < Ultra
Just Apple thing, Maximum does not mean Maximum.
I don't think Apple has ever officially said Max = Maximum. Could be someone's name :) The bigger confusion to me is it's pronounced Macs!
If the difference between M6 and M5 Pro or M5 Max is confusing to you, I strongly recommend you don't look at the names of literally any other CPU, GPU, or chipset that is being manufactured.
Yeah, but traditionally Apple doesn't pull that weird naming shit. Although I guess the mobile naming scheme has been headed there for a while.
they used to have 1 iphone, now there are 6 models with around 21 exterior versions total
And you think 1 is better.
I think it's simpler
Not better - but Apple in early days was easy: Choose Air if you don't nedd much. Choose MacBook if you need a little but more and choose Pro if you need max power. That's one of the reasons why I bought an Apple years ago.
Compare that with the regilar lineup of other manufacturers: Cryptic model names which are uncomparable and models that are so similar that they could be equivalent.
So, no, 1 is not better but I don't care (as customer) about variety.
Well that was sort of the case. You had different years of each model to consider as well, which isn’t all that different from the M/Max/Pro thing we’re discussing here. They didn’t refresh all products the same year.
> And you think 1 is better.
I think the Apple experience was better when they focused on delivering a single SKU of each product line, yes. Less decisions as to what to buy, greater focus on making that one SKU stand out.
Are you kidding? When have they not? Apple II, II+, IIc, IIc Plus, IIe (also //e), IIe enhanced, IIGS. All sold concurrently.
They’ve always been this way.
Most of those models were not sold concurrently. The Apple II, II+, and IIe each replaced the previous model; the same was true of the IIc and IIc Plus. "//e" was an alternate textual rendering of IIe, not a separate model.
The most Apple II models that were sold concurrently were the IIe Platinum, IIc Plus, and IIgs. (And this was around 1990, at which point there were far more models of Macintosh computers on the market.)
//e was a separate model from the ][+
64kb RAM + 80col card vs 48kb/40col
The 2 and 2+ were concurrent.
The 2e, 2e enhanced, 2c, 2c+ and 2gs were concurrent.
I didn’t bother with Mac because I got tired of writing all the things.
Bad example, but very true with the Mac under sculley. When Jobs came back that was one of the things he greatly simplified
The second iPhone was the 3g the third was the 3gs.
The second one was the first iPhone with 3G networking (which had broadly been considered a required feature for smartphones when the original iPhone was introduced without it.)
I won't try to defend the 3Gs ;)
At least the N, Ns naming scheme referred to straight-forward (albeit minor) upgrades, and were released a year apart...
Not to mention X and Xs.
And the fact that they skipped 9 and will likely skip iPhone 19 as well
Microsoft skipped Window 9 too.
Samsung skipped an entire range of numbers from Galaxy S 11 to 19.
Looks like tech companies don't like 9...
Supposedly Microsoft skipped 9 because of a lot of hardcoded checks for windows 95 and 98 that checked the first digit
And the iPhone XR! All part of the twelfth-gen models. Of course the subsequent thirteenth-gen models were just called "iPhone 11" in a return to sanity.
I still don't remember what X vs XR vs XS meant. Of course I can look it up, but that's not the point.
Tell me you're not old enough to remember the Performa and Centris lines without telling me you're not old enough to remember the Performa and Centris lines.
All the people under you not understanding that you meant AMD repackaging 2-4 year old CPU cores as """new""" by giving them a current series name and bolting some extra features on.
> I strongly recommend you don't look at the names of literally any other CPU, GPU, or chipset that is being manufactured.
And certainly don't look at MikroTik's product naming. ;)
Put another way, it is a peculiar naming regime when M(n-[2-4]) Pro/Max/Ultra is superior to M(n). This is very visible right now because of the demand for running local AI on Macs.
It's way better than the naming scehme of Ryzen and Intel Core i series.
Threadripper’s model numbers where briefly their msrp.
It's also kinda sad that there's no clear-cut best chip, e.g. the M6 outperforms the M5 Pro in some ways (singlethreaded performance, maybe NPU and stuff) but then the M5 Pro also outperforms the M6 in some ways (multithreading, graphics). This was even worse when the Mac Studio was stuck on M3 Ultra for a while when the M5 Max was out. The M5 Max was simply better at most things and it was ironic that the fastest computer you could buy was briefly a laptop.
How could it possibly be otherwise? Single core performance goes up every year, but not enough that this year’s 12-core CPU beats last year’s 18-core CPU at multithreaded benchmarks.
This is “sad”?!? First world problems.
It's also not exclusive to Apple; the normal product cadence from Intel and AMD usually has the large workstation and server chips falling behind the consumer chips in peak single-core performance for part of the product cycle. Some market segments just don't sustain annual updates very well, and it's always easier to get a smaller chip out the door first before a larger chip.
In a world with the chips themselves named things like:
Ryzen AI Max+ Pro 495 or Snapdragon 8 Elite Extreme Gen 6.
Your distaste seems misplaced.
> Ryzen AI Max+ Pro 495 or Snapdragon 8 Elite Extreme Gen 6.
Don't even get me started on USB...
They're all bad, some are worse
We should have enough distaste for them all
A name like M5 Pro denotes the generation and the number of cores.
What does a name like Snapdragon 8 Elite Extreme Gen 6 even mean?
You can just Google it. It's not a complex answer:
- Snapdragon 8 is the highest-end mobile SOC line, as opposed to midrange Snapdragon 7 and low-end Snapdragon 4.
- Elite means that the CPU core is Snapdragon Oryon, and not Arm Cortex off-the-shelf designs.
- Extreme is a trim/upgrade package that gives the GPU more power and performance headroom.
- Gen 6 tells you the generation of chip from that line.
You should refer to prior discussions on this topic if you're genuinely confused, or somehow think that Apple has a superior taxonomy: https://news.ycombinator.com/item?id=50116504
Latest vs. greatest. Buyers who’ll pay for performance can deal with that discomfort and be squeezed for buying the best, but those who need an extra push to buy Apple might be sold on the idea of the latest.
bigger number = more megahertz
This sounds exactly like the total confusion of people looking at Luna 6, Terra 6.1, and Sol 6.1.
I genuinely do not understand what is "confusing" about this. Like, at all. Annoying maybe? But what is confusing?
> Phenomenal battery life: In addition to the huge leap in performance, battery life gets up to 14 additional hours for Intel-based upgraders
the m2/m3 announcements were already kind of cringe to keep saying 'HEY IT'S BETTER THAN INTEL' but at this point it's just embarrassing?
I would guess there's still quite a lot of people on Intel Macs out there (they were sold 5 years ago), so it is still very much worth it for Apple to market to them.
Why? Apple devices have the best battery life in the market right now to my knowledge. It's an important part of the evaluation for a laptop which is the opposite of cringe.
Actually, they don’t! Intel Core Ultra Series 3 beats Apple laptops as of the M5 on battery life.
IIRC the Framework 13 Pro beats the MacBook Pro at both video streaming tests and office productivity with 300 bits brightness tests. (Framework live-streamed the video playback test, and the office provided productivity tests I referred to were done by Just Josh Tech)
Perhaps the M6 has leapfrogged but that remains to be seen.
Competition is a beautiful thing.
In usual developer type tasks, among my friend group what I hear is that panther lake laptops are competitive with M4 laptops. But, M4s do that with a 55Wh battery, whereas most panther lake laptops come with a 70Wh battery. So if you're discussing chip performance per watt then it's probably an edge to apple silicon.
72.4 Wh is the size of the M4 MacBook Pro battery, about the same size as the Framework 13 Pro.
Intel and now Qualcomm are really not far off on performance per watt. Mostly Apple maintains an advantage on single core performance.
The Panther Lake integrated GPU in the X7/X9 SKI is actually faster than the one you get in the base M5.
You're right. I think I remember those numbers from an Air vs 3xxH comparison.
Neither of those have comparable perf
They do, please look up reviews of Panther Lake 338H/358H. Example: Lenovo Thinkpad X9 15p Aura Edition
I did. I am seeing single core geekbench scores of 2700 compared to my MacBook Air at 3500. This, for a laptop that is thicker and heavier than my air.
why tack on the goalpost moving? And geekbench single core isn't the only measure of performance.
in my original post, i said "perf", i am not sure what you imagined, but i am not moving anything, geekbench single core score is a perfectly reasonable thing to measure under the name "perf"
1. It’s Geekbench, not a real experience of using the system. A geekbench number being xx% slower doesn’t mean that the system is xx% slower at doing tasks. It’s a synthetic benchmark.
2. The Intel chips beat out the M5 in multi-core and GPU.
3. These Geekbench scores are close enough that they are in the same performance class. We aren’t talking about 1,000 versus 3,500. The end user isn’t noticing a performance difference between 2700 and 3500, especially since that number is single core only (most workloads can take advantage of multiple threads these days).
Even if we concede that the Intel chip is slower for single core tasks only we can’t just lie and say it’s not in the same performance class. It’s very clearly comparable by objective data.
What we can say definitively by objective measures is that Intel is behind by a year at most, which essentially doesn’t matter in the grand scheme of “which laptop to I buy?” It’s close enough that the processor shouldn’t be part of the buying consideration for the majority of consumers. Intel is not years and years behind what Apple is putting out there, this is not the M1 where Apple was beating Intel in performance and getting double the battery life. We are looking at much more incremental differences where the buyer will not notice a difference over the lifespan of the product.
multi core scores? So you'll happily accept a 100,000-core pentium one class CPU to replace your laptop? it will have higher multicore score. No? Really? Oh? why not?
What are you talking about? Is this supposed to be some kind of gotcha?
Here’s a fun one: look up Cyberpunk 2077 benchmarks on Panther Lake compared to M5. It runs natively on both Windows and Mac.
That's not what the marketing is saying, they keep comparing to their own super old intel models.
I recall seeing some crazy stats like 41+ hours maybe on a snapdragon processor w/windows. I imagine your screen brightness would have to be like 99% dark and not playing videos is my guess.
On the other hand their hardware only supports one operating system.
So it's not an apples-to-apples comparison.
More people than you think (outside tech) are using Intel Macs at the moment (family computers etc).
Exactly. A late model Intel Mac is still a perfectly cromulent computer in 2026.
"cromulent" sounds like such a disgusting word that isn't.
probably because of its adjacency to "corpulent" lol
The word "cromulent" was coined by the TV show The Simpsons in a gag.
> Mrs. Krabappel: "Embiggens"? I never heard that word before I moved to Springfield.
> Ms. Hoover: I don't know why. It's a perfectly cromulent word.
But now it has made it into the Merriam-Webster dictionary.
There's no happy medium here that would satisfy everybody: when they make marketing claims about improvements relative to products from several years ago, they get criticized for not focusing on the improvements relative to last gen (even if those comparisons are also somewhere in the announcement). When they focus on the most recent generations, people complain that it doesn't make sense to upgrade every year.
Just accept that even if you aren't still using an Intel MacBook Pro, the comparison is still relevant to some part of their target market.
They only just dropped support for Intel Macs from this fall's operating system release (but are still doing security patches for the releases that Intel Macs can run). A year from now, it actually will be fairly reasonable to expect them to give up marketing to the holdouts.
Fun fact: the laptop in this size/performance class with the best battery life is a Windows Intel Core Ultra Series 3 laptop, not a MacBook Pro!
So Apple is now comparing against computers they were selling 7-8 years ago rather than what Intel is delivering now.
> Apple is now comparing against computers they were selling 7-8 years ago
Apple didn't fully stop selling Intel macs until June 2023. And they compare against the most recent previous model too.
in 2023 they were only selling desktop Intel Macs. I am pretty sure by 2021 they weren't selling Intel laptops unless maybe refurbished
> Fun fact: the laptop in this size/performance class with the best battery life is a Windows Intel Core Ultra Series 3 laptop, not a MacBook Pro!
Source for that fun fact?
I think they refer to low-tier Wildcat which, indeed, is more efficient than Apple's CPU in Neo, which they typically get compared against.
Nope! I’m talking about the 358H compared to the M5. Multiple similar Windows laptops outlast the M5 MacBook Pro and Air.
Framework 13 Pro, Lenovo ThinkPad X9 15p Aura Edition, for example.
Have you considered battery in them?
Just Josh Tech’s review of the Framework 13 Pro or the Lenovo ThinkPad X9 15p Aura Edition. Multiple similar sized/classed laptops get better battery life.
This is of course as of the M5, M6 hasn’t been tested yet.
last time I checked, every Windows laptop had the absolutely massive caveat that it needed to throttle performance by a ton to have a low enough TDP to compete on battery life. You just cannot match the efficiency of ARM Apple chips with an x86 chip.
But the problem is “last you checked” is before Panther Lake came out.
It is not 2020 anymore. Look up reviews. Just Josh Tech’s ThinkPad X9 15p Aura Edition review, for example.
What you’re saying is just absolutely not true at all and is only backed by your gut feeling.
As far as I can tell after just looking at the specs, what I said is still the same. For example, an Intel Core Ultra 5 Processor 338H has a TDP of 25-65W vs. the M5 TDP of 3-27W. These aren't even directly comparable, but it gets worse going higher in the Intel Stack.
Further, every Windows laptop seems to have a worse display. For example, an M5 Macbook Pro 14" has a 120Hz 254ppi mini-LED display with a peak brightness of 1600 nits vs. a Dell XPS 14" with a 120Hz 161ppi LCD display with a peak brightness of 500 nits.
So it seems like they have figured out that they can put bigger batteries, worse displays, and throttle TDP to get better battery life, but it is still not possible for them to beat Apple's efficiency just based on those TDP numbers alone—these chips are just drawing more than 2x the power at peak load.
I don't understand why there isn't more effort towards ARM in the Windows world.
They want those users to upgrade, I don’t see the problem.
I think it makes sense. Do you think most potential buyers are coming from last year's model, or from laptops that are more like 3-5 years old?
As long as they're clear about the comparisons they're making ("X% more battery life than Y" as opposed to simply ("X% more battery life") I'm good with it.
I’ll wait for it to surface on eBay in few years and then consider again. Pretty sure many of these go in the hands of people who can never make good use of all the nice hardware in it.
Shame they are not releasing M6 pro/max on a 16 inch, was a good time for me to switch from M1!
I always max out whenever I upgrade my MacBook Pro.
Last time, in 2024, it cost me $6K
Today, it’s $10K
Yowza
I've always done the opposite; getting the base model (maybe bump 1 spec), but then just stay current.
Also here: https://news.ycombinator.com/item?id=50258521
dupe https://news.ycombinator.com/item?id=50258521
I have a 14” M2 that you’d need to prize from my cold dead hands. I can’t imagine buying an M6. There just isn’t any need.
I am due to an upgrade (still using the 2019 intel). However I am torn between a M5 Pro with 48 Gigs of ram or waiting for M6 pro. I’d really like to run local models for once.
won't be an M6 pro, they are saving the ram and other components for the M7 series
Interesting. I guess I’ll have to go with the m5. On one hand it feels bad to buy older generation hardware. But then. It’s already so much faster than what I have…
Still using my 8gig m1 to make mac builds, its slow, but works fine. Keep thinking about upgrading, but max 32gigs ram, if I can hold out for 2 years, and maybe have 64/128gig as options, and cheaper.
In your shoes, I'd just get whatever Pro processor is available at that time. (Knowing that new processors were coming, I happily bought an M5 Pro machine last month)
32GB Unified RAM max. Is that really the machine I want for running local LLMs?
Nope.
This M6 is the base Pro model. You probably want the M5 pro/max for your LLMs.
It's called Pro, but it's hardly Pro, more like a MacBook with a good fan.
Not really - it seems that M7 will come out in Q1 27 and be targeted at that use-case more specifically.
Base M7 probably will still have pretty low memory bandwidth for serious AI workflows. M7 max/ultra might be pretty interesting though.
Hard for me to get excited about these anymore.
Apple software is getting worse every day and while they're not Google or Microsoft they're getting more user hostile at every turn as well.
The lengths they go through to prevent you from doing what you want on their (not your) devices is astonishing.
Very happy with my framework running fedora for many years now... Sure it's not an M1000 but it's almost as good and properly mine
[dupe] https://news.ycombinator.com/item?id=50258521
Where's the touch screen??? The rumors promised touch!!!
Why would you want greasy fingerprints all over your display and gorilla arm syndrome when there's a perfectly fine keyboard and touchpad attached to the display for free?
mmmmm....grease
No reason for gorilla-arm when touch is an alternative interaction mode and not a required one. You still have trackpad and keyboard. You can dynamically switch modes depending on the action.
Gorilla-arm was a fear based on the naive idea that if you had touch, you would be required to only use touch for every action.
Is apple going to be so restrained to keep the design of the OS as a mouse first experience or are they going to cripple information density so I might be able to close a window with my elbow?
People act as if Chromebooks with touch screens aren't already considered a huge UX upgrade over Chromebooks without. Use your finger where it makes sense and the mouse where it doesn't. Literally nobody gets tripped up over this in practice.
I have a surface laptop with touchscreen, and i can on my hand how many times i have used it this year. Mostly to check if it's still working.
Never understand the appeal of touchscreen on laptop.
Chromebooks run Android apps natively, so you can flip the screen around and use them as pure tablets or just use tablet apps in laptop mode just fine. For math, you can just write directly on the screen. In schools, kids playing Blooket/Gimkit/Kahoot on touchscreen devices crush those using a touchpad.
I don't actually want a MacBook at all, I was just joking.
This is only the base model MBP. The high-end touch one will be later this week.
the rumors were about a supposed Macbook Ultra that would have a screen like the ipad pros. Still to come (or not)
I would pay 2x or even 3x for this with working Asahi Linux patches provided in source code form on Apple's site.
doubt
Asahi supports up to the M3 line. Linux 7.4 supports M4 so surely Asahi will support it soon.
What is it about this MBP that makes it so much more preferable to an MBP with M3 Pro/Max?
I want the newest one and I want to run Gentoo Linux on it with working Asahi patches for all aspects of the hardware. I want it enough I'm willing to pay for it. I'll pay 3x Apple's current sticker price.
But Apple will never sell that to me, as far as I can guess. Apple is not a good person.
And I want a pony
> Apple is not a good person
It's a company.
Nice, I'll keep my 64GB m1.max though.
I love that the house is on fire and Apple wants to drone on about "battery life". They have always been a company in denial but wow.
Still using the M1 Macbook Air for 4K video editing on Davinci Resolve and Photo Editing with Darktable (open source). Just 8GB of Ram too and I've never really needed more. Obviously these new versions are great but even the original is not too shabby.
i hope they're not launching that touch stuff... just stick with these pls
The design is really great for the new MacBook Pro. The cpu is no compare to other laptops.
One thing, framing AI use for the base M models marketing makes no sense to me. But honestly this computer is still faster than any of my laptops that came before, even with the visual atrocity of Liquid Glass.
I'm currently having an M2, but I just can't imagine buying an Apple laptop again. The hardware is great, but the sofware is just unbearably bad. I used Arch Linux with KDE before, and it was just an experience 10 times better. If they would just support Linux out of the box...
I would not get this because I want to run LLM's locally, you are really gonna get gpt 4 performance and constantly looking at the SOTA or the open models that needs way more to have it work practically, and without your laptop sounding and feel like a hair dryer.
Did they figure out how to fit an LTE modem into it?
Anyone designing their own macbook clone using llms yet?
This is going to be great with Asahi
I need 64GB RAM and 1TB SSD suddenly having been more than comfortable with 16GB and 256GB respectively. My agents have sessions which have workteees and scratchpads and simulators and build runs and test suites and wow. The hours I’ve been spending for memory swaps and quad simulator calls to run through. I am compelled to view a $4k machine as a necessity. Also a cool thing about M5 Pro is that it’s very close to Max on the CPU performance side with identical core count if you get the 18 core variant. The Pro’s Geekbench scores are also in similar ballpark as the Max. You rather get the Pro for most use cases and use the extra cash on RAM.
But can it run Omarchy :)
The "notch" seems to be gone based on the product pictures. Time to upgrade from my ~2014 model MBP.
I do see it in one photo on their site. I just hide mine with software ("topnotch") and it's fine.
Looks just as notchy as they've always been.
One of the interesting slights of hand Apple marketing must do is simultaneously say the new processor is awesome while selling the higher spec versions of the previous processor as upgrades:
sleight*
My 5090 has a bandwidth of 1,792 GB/s, and that is just about good enough for basic (i.e. 27B models) local LLM inference.
The fastest M6 MacBook Pro (24/32GB configs) has a memory bandwidth of 170 GB/s. And the M5 Max is 614 GB/s.
Am I missing some nuance of the M6 MBP here, or am I right to assume it won't set the local LLM world on fire?
Is your 5090 in a laptop? o_0
No. But does high memory bandwidth necessarily require desktop/server style power and cooling?
Laptop 5090 (24gb) is 800gb/s.
M5 Max (128gb) is 600gb/s
Rtx spark laptop (128gb) is 200gb/s.
Desktop 5090 (32gb) is 1700gb/s.
Desktop 6000 (96gb) is 1700gb/s and 600W.
So, yes?
The case of your 5090 having become a coveted artifact of great rarity in these times - may be a nuance of some measure.
Also, the following two have very different 'needs' to feel good enough:
1. heavy agentic workloads via harness vs 2. a simple private chat interface with a local instance
Even if an MBP's bandwidth were higher, it's tough to see a reason to run it as a constantly loaded 24/7 inference machine due to heat degradation.
P.S. still extremely awesome to be able to host and chat with a decently quantized 27B model on your laptop. ...and to be fair even a 32GB Air would do that! It is a thing of beauty: the amount of folks this may add to the hacker/tinker community.
Isn't the laptop version of the 5090 is a more apples-to-apples comparison? It has 896 GB/s memory bandwidth (compared to the 614 on the M5 Max MBP).
The 5090 does have a bandwidth advantage, but also must pass more data from main to GPU memory for any particular task. Comparisons of a 5090 to a DGX Spark at 273 GB/s are illustrative since the Spark uses unified memory similar to Apple's M-series. [0] The 5090 certainly wins most benchmarks but that doesn't mean the DGX Spark is not useful.
0. https://float16.cloud/en-en/ai-benchmark/dgx-spark-vs-rtx-50...
Loading data to the GPU will generally never be the bottleneck. Yes that's a bit better for unified memory, but prefill and decode will still be bound by compute and memory bandwidth respectively.
There's some value in having a huge amount of memory addressable by the GPU for loading larger models, but even that doesn't make sense for current gen models. For MOE models, newer inference engines with smarter GPU expert caching and pinned memory allow a GPU with lots of bandwidth to significantly outperform a Spark, so that leaves you with larger dense models that fit on a spark but not on a GPU where you might have a use-case.
But even then it's hard. For instance a full-fat BF16 of a ~30B parameter model will fit on a 128GB spark and not a 32GB 5090, but at that size it's going to be frustratingly slow on the spark given the compute and bandwidth constraints.
Imo these devices are mostly good for experimentation, or if you only want to do overnight jobs or training which wouldn't be possible on a GPU, and you don't want to pay $100k for a DGX station.
too bad only the iMac gets all the color!!!
Dare I even ask how much the 32gb/1tb variant will be? $3,000?
You could have dared to check. But no, it's $2399 for those specs.
does it run linux?
Does it run Linux?
I can't understand this people
Windows on Mac, good times ;)
M6 Pro wen
It's not justifying an upgrade. Shame there are no finance options for M5 Max 128GB.
About time they put a battery in the lid, which is what an iPad already manages to do with thinness. Boom another half a decade of updates all the way to 48hrs
Citation?
It's confusingly worded but I think it's a suggestion for something Apple could do in the future. I don't see any mention of this in the linked Apple article.
Also a great way to make money from screen replacements when the battery expands!
I think the 16" pro is already at the max battery size you can take on a plane.
i wonder if it has 2 fans
Apple can release as much fantastic hardware as they want - as long as the software keeps getting worse, what's the point?
eh no thanks. not jumping to m5 max now we know apple is cooking with m7 pro and m7 max. don't drop the ball.
you know they jut bump this every year right
The rumour is they aren't going to bump the pro and pro max.
Yes but have you considered the M8?
Imagine, these are the machines we get to vibecode on.
Been waiting to get a new MacBook and that day has finally come! Thanks for sharing