Second IC – Homemade 1000 transistor array chip (2021) | Hacker News Reader

Second IC – Homemade 1000 transistor array chip (2021)

sam.zeloof.xyz

271 pointsdanboarder37 comments

And now, a couple years later, he's partnered with microprocessor legend Jim Keller to start Atomic Semi.. https://twitter.com/szeloof/status/1636174775310745601

I didn't make the connection until now, but Jim Keller was on the Lex Fridman podcast twice [1, 2] and it was awesome. In terms of founders this is pretty great team.

[1] https://www.youtube.com/watch?v=Nb2tebYAaOA

[2] https://www.youtube.com/watch?v=G4hL5Om4IJ4

I love the audacity of this as a hobby project. It makes me wonder what other audacious hobby ideas would be dismissed by almost everyone but are actually achievable by a suitably motivated individual.

Have been thinking about making an authentic copy of V2 rocket since like 20...

Just don't launch it at authentic targets... :D

There's a whole sub-genre here of projects that would be (wisely) dismissed because they would land you in jail rather than their technical difficulty I suppose. I recall recently seeing a rocketry hobbyist realize their landing guidance tech is essentially the same as precision missile guidance and best not to make open source.

EDIT: I just remembered, it was Mark Rober's egg drop from space [1]

[1] https://youtu.be/BYVZh5kqaFg

Tangentially related, it appears the US weapons industry uses hobbyist parts to build drones. https://en.defence-ua.com/news/russian_soldiers_disassemble_...:

“A video appeared on the network shows russian soldiers disassembling the American-made Switchblade 600 suicide drone manufactured by AeroVironment. According to Bild, it is the first documented use of this drone by the Armed Forces of Ukraine.

Moreover, the drone is equipped with 3 “aero-naut CAM Carbon” folding propellers produced by German company Aero-naut Modellbau. It’s a German company that specializes in designing and manufacturing model airplane and boat kits, as well as model components such as electric motors, speed controllers, and propellers. But the company has stated that it never intended for its products to be used for military purposes, and has expressed concern over the use of its products in military applications without its knowledge or consent.”

I know a guy that made a ramjet, so that's definitely doable.

If you really want to go retro you will have to commit to invading Poland...

A valveless pulse jet is on my bucket list of things to build. First an electric cargo bike after finishing my garage.

What about making everything else (all the tools) that Sam uses to make the silicon chips? How would you make one of those Nikon projectors? The test equipment? The chemicals? Like some comprehensive post apocalyptic bootstrap.

You might like this: https://www.thomasthwaites.com/the-toaster-project/

Achievable if someone bankrolls you sure

> but are actually achievable by a suitably motivated _and minted_ individual.

Unfortunately, in my country something like this is impossible for most people, simply because they don't have space and money to do anything.

I had some many projects in mind when I was younger, but my parents were poor and I had a tiny room. We had no garage, no garden. I still saved money to buy soldering iron and electronic components to experiment with, but I was damaging them at such a rate I spent all my savings on components and had to abandon this for many years.

I still cannot afford to rent a place where I could do something that interests me and I have a good job. Things also have become much more expensive, so it takes forever. Like I can only afford to order one iteration of project per month and it's frustrating I have so many ideas I just don't have anywhere to make them.

I came up with a system, so that I can store everything in boxes and have stacks of them from the floor to ceiling and a spreadsheet where everything is located, but it takes so much time to take something from the bottom or when I put breadboards in and accidentally disconnect something.

Sometimes I get so depressed by this I can't even do anything for weeks.

> Unfortunately, in my country something like this is impossible for most people, simply because they don't have space and money to do anything.

Why is the country so unproductive?

Not every country is in the position to just print money, you know?

No country is in the position to print money -without causing massive inflation. It's a common myth that the "west can just print money".

The USD is the key currency. If the US prints money it is a different thing than e.g. if Denmark prints it's kroner in terms of financial consequences.

Not really.

A currency's value is determined by how confident people are that it will hold it's value. It's driven by supply and demand.

If country A trades with country B, it'll either get paid or pay with country B's currency. Country A might not be interested in country B's currency, so it can use the US dollar and a US institution as a proxy between the two. But that doesn't do anything to affect the value of the three currencies. It's simply a matter of convenience.

What matters is how confident people are of their respective government, laws and institutions compared to the US. Would they prefer to hold assets in the United States or their own country, or country A or B.

You should look up for hackerspace's in you city, it's a space with tools to do projects.

Years ago I tried, but these places are not friendly to neurodivergent people. Usually bullies take all the space and if they see you do something interesting they hang around to steal whatever you are doing. There is nothing you can do about it if you have no money.

Sorry to hear you had this experience, I would say its worth giving another go maybe you can check out the IRC first to check the vibe before committing.

It very much depends on which one, I find - the local one is primarily run -by- neurodivergent people so that tends not to be a problem.

templeOS comes to mind as an audacious hobby project. Even if it was more mental illness than audacity.

I don't know how much commercial value his projects may have, but I really think they could be used in education. It very is rare for students (even in developed countries I think) to be able to develop and test their own chips.

Having a simple, practical, affordable "chip making" kit that any one could have access would be a boost for students all over the world.

I don't think any of these adjectives applies.

Amazing work. Would be great to see more of the gear used and the processes themselves and how feasible it is to reproduce this. I'd absolutely love to be able to do for instance a homebrew 6502 or so. But I fear the gear costs will be astronomical.

Check the 'general high vacuum' page linked on the left there as well.

A home-made lithographically-fabricated integrated circuit (2018) - https://news.ycombinator.com/item?id=28263714 - Aug 2021 (13 comments)

Second IC - https://news.ycombinator.com/item?id=28176255 - Aug 2021 (79 comments)

Homemade IC (2018) - https://news.ycombinator.com/item?id=21239145 - Oct 2019 (40 comments)

Home Chip Fab - https://news.ycombinator.com/item?id=20657398 - Aug 2019 (131 comments)

A home-made lithographically-fabricated integrated circuit - https://news.ycombinator.com/item?id=16955549 - April 2018 (116 comments)

Loved reading his blog over the years. Another good resource in this area is the zero to asic course:

https://zerotoasiccourse.com/

Been following Matt Venn of zero to asic for a while. Hes got a great youtube channel as well. Looking forwards for when they can support designs a couple order of magnitudes larger. I really want to design an audio spectrum based OFDM modem 802.11a-ish. Might be too big of a design for the foreseeable future tho.

This is really fantastic work and there definitely needs to be a hacker microchip community. I love the idea of one day having a simplistic processor chip someone made in their garage.

This kind of real passion deserves much more representation in the news.

May Atomic Semi be fine. May the God of Engineering bless them.

That Industry vs Sam's garage chart is hilarious :D Obligatory XKCD: https://xkcd.com/605/

Yes, at this rate he'll outrun TSMC in a couple of years.

Archived version: https://archive.is/SuAf2

This is so incredibly cool. I've always wondered if this were possible and it tickles me to see it done - starting in high school, no doubt.

Sam's garage looks like a billion dollar lab, pretty affordable for a DIY.

Reply on news.ycombinator.com