First thought when reading the headline was that the "Man" I would expect to do such a thing would be Sam Zeloof. If it had said "Woman" instead, my guess would have been Jeri Ellsworth ;)
"By switching from a metal gate process to a polysilicon gate process, he was able to reduce the chip's required voltage input from 10v to just 1v, reducing power consumption significantly." - doesn't power consumption of chips scale with U^2, i.e. reducing voltage to 1/10 makes power consumption drop to 1/100? I could not find the numbers yet...
I showed this to a friend who works at one of the big chip design companies. He thought it was cool, but pointed out some limitations of the current setup.
1. He's going to need multiple device types. My friend discussed the requirements for various logic families (cMOS, nMOS, DTL, and RTL). RTL would probably be the easiest.
2. He needs interconnect.
He said that process-wise, this is about where the industry was in the mid-70s, and that fab lines weren't super expensive then, in the 10s to low 100s of k range. With used equipment and improved technology it would cost less to run a similar process, so what the guy is doing is pretty cool and he's got room to grow.
So...the guy is still a long way from a 4004, but he could maybe get there.
Also the title is misleading. It says it's "Similar to Intel's 4004 CPU", but the similarity is only in transistor count (and even that's off by a factor of 2), not complexity/functionality. Sam Zeloof's IC is basically a grid of non-connected transistors, whereas a CPU is a bunch of transistors arranged in such a way to allow for computation.
I came across him on Twitter while back and found his work pretty fucking cool.
Only thing I haven’t bothered to figure out is how he managed to get started in all this. Surely doing homefab requires equipment that costs a lot of money? I can’t imagine less than tens of thousands of dollars even for very old equipment.
I would expect that it is certainly expensive but he seems to be building lots of his equipment himself instead of buying it. The photolithography machine is a modified microscope with a projector for example. The centrifuge for spin-coating the wafers with photoresist is also DIY'ed according to the video from the article.
First thought when reading the headline was that the "Man" I would expect to do such a thing would be Sam Zeloof. If it had said "Woman" instead, my guess would have been Jeri Ellsworth ;)
"By switching from a metal gate process to a polysilicon gate process, he was able to reduce the chip's required voltage input from 10v to just 1v, reducing power consumption significantly." - doesn't power consumption of chips scale with U^2, i.e. reducing voltage to 1/10 makes power consumption drop to 1/100? I could not find the numbers yet...
The "Industry vs Sam's garage" chart is pretty funny.
I showed this to a friend who works at one of the big chip design companies. He thought it was cool, but pointed out some limitations of the current setup.
1. He's going to need multiple device types. My friend discussed the requirements for various logic families (cMOS, nMOS, DTL, and RTL). RTL would probably be the easiest.
2. He needs interconnect.
He said that process-wise, this is about where the industry was in the mid-70s, and that fab lines weren't super expensive then, in the 10s to low 100s of k range. With used equipment and improved technology it would cost less to run a similar process, so what the guy is doing is pretty cool and he's got room to grow.
So...the guy is still a long way from a 4004, but he could maybe get there.
blogspam/dupe of https://news.ycombinator.com/item?id=28176255
Also the title is misleading. It says it's "Similar to Intel's 4004 CPU", but the similarity is only in transistor count (and even that's off by a factor of 2), not complexity/functionality. Sam Zeloof's IC is basically a grid of non-connected transistors, whereas a CPU is a bunch of transistors arranged in such a way to allow for computation.
I had a feeling this was about Sam Zeloof.
I came across him on Twitter while back and found his work pretty fucking cool. Only thing I haven’t bothered to figure out is how he managed to get started in all this. Surely doing homefab requires equipment that costs a lot of money? I can’t imagine less than tens of thousands of dollars even for very old equipment.
I would expect that it is certainly expensive but he seems to be building lots of his equipment himself instead of buying it. The photolithography machine is a modified microscope with a projector for example. The centrifuge for spin-coating the wafers with photoresist is also DIY'ed according to the video from the article.
http://sam.zeloof.xyz/maskless-photolithography/