A home-made lithographically-fabricated integrated circuit (2018) | Hacker News Reader

A home-made lithographically-fabricated integrated circuit (2018)

sam.zeloof.xyz

114 pointspunnerud13 comments

Sam's second IC, the Z2, was discussed a few days ago. You can find the discussion here:

https://news.ycombinator.com/item?id=28176255

dang has a comment with links to other related discussions in there.

Thank you, did not know it was posted before

If you are curious how Sam is creating these chips, he has a video [0] about his maskless [1] photolithography machine here. He built it ("cobbled it together", in his words) from a microscope and a DLP projector. The microscope is used in reverse, so he's feeding the image into the wrong end (i.e. where the ocular would be) and uses the optics to concentrate it onto the wafer. The DLP was modified by removing the color wheel because that attenuated the UV radiation that is wanted to cure the photoresist on the wafer. A lot of his equipment is home-built, which, as I would guess, saves him a lot of money.

[0] https://www.youtube.com/watch?v=Nxz_ENnmgtI

[1] A photomask is a plate with the patterns of the features of the design that will be projected onto the wafer, see: https://en.wikipedia.org/wiki/Photomask

It took years for this man to get to the point where he is now, and all this with modern equipment, and widely available literature.

Imagine what a monumental feat it was 50 years ago.

Him doing it all by himself is something quite extraordinary by itself. I can't imagine if that all can be replicated even by a mid-sized research university with decent labs (academia, and research outsource almost everything from their lab scale chip plants besides things which make their area of expertise.)

> a mid-sized research university with decent labs

What kind of university are you talking about? This could be anywhere from some random state school to Princeton.

A university which can afford semi equipment. Even 30 years old process tools cost eye watering sums.

> cost eye watering sums

The good universities like Princeton/MIT/Harvard have literal billions, so semi equipment could be a drop in the bucket for them.

Not every university is Princeton/MIT/Harvard

Well, maybe they could share access with some less endowed institutions.

You know what they say, go big or go home

From a theoretical perspective, the main achievement was the junction diode. Everything after that was incremental.

Not quite 50 years, but in 1982 under the Iron Curtain two Polish students developed Intel 8080 clone as a PHD thesis. Manufactured between 1982-1993. http://retrokolekcja.pl/MCY7880.php

For comparison original 8080A: http://retrokolekcja.pl/zdjecia/8080A%20struktura_d.jpg

Polish MCY7880: http://retrokolekcja.pl/zdjecia/MCY7880%20struktura%20ITE.jp...

Russian KR580BM80: http://retrokolekcja.pl/zdjecia/MCY7880_zsrr.jpg

And the absolutely amazing gem. In 1983 legendary Polish science education TV program SONDA documented design and manufacturing of first batches in a humorous lets bake a cake fashion. Paper plotters, light pens, developing/rinsing dies by hand, electron microscope debugging, the whole nine yards!

part 1 https://www.youtube.com/watch?v=AJGp7keIA_o

part 2 https://www.youtube.com/watch?v=KHl6m93Hay0

part 3 https://www.youtube.com/watch?v=rcOTwkT-PDU

Amazing anything worked considering really Spartan conditions. Yields were low and political pressure from Moscow meant later batches were mostly KR580BM80 dies packaged in Poland.

wow didn't realise just how small they are until the pic with the screw in the picture showed up. Impressive work!

Reply on news.ycombinator.com