Open Source Integrated Circuit Manufacturing

[Alexander S.]

If anyone is interested in collaborating, i'm currently researching integrated circuit manufacturing in the context of this ‘commons based peer production’ initiative.

The goal would ultimately be to design a set of open manufacturing systems (e.g high-purity silicon production, oxidation, photo-lithography…), to directly transform locally sourced waste and raw materials into ICs on much smaller scale and throughput then current IC fabs. Additionally, adapting each system and material choice to be non-ecologically destructive (or zero waste), as well as ensuring the local development of the back end manufacturing processes that can produce the machines that fabricate integrated circuits (Back end processes > IC producing machines > ICs).

Any thoughts/insights/criticism would be greatly appreciated, or if anyone is interested and would like to embark on this let me know. 

[Anthony Douglas]

I would.  We need a facility, and some colleagues.  We could use better upstream production equipment, some stuff better than three d printers.   The facility will have to include housing for the humans, a machine shop. A well stocked parts library.  A variety of materials like chemicals and stock materials.  A bunch of money to buy mor stuff.  Some computers, of course.

Enoch freedom com regulators and insurance people to get on with things.

Then, some time.  A year, maybe, with what 20 people, without distractions or impending doom, might be enough to deliver a small garage scale system to go from commodity  inputs to some FPGAs that are reasonably useful? I font know if that's enough.

I'm surge we can find some volunteers, but really there abound be some money in it.  We could try crowd funding.  But I ultimately have to admit there is no major pressing need for this ability to produce ics at a small scale, and that seems like it would undermine funding.  

Even if produce such a system, or doesn't strike me as a panacea.  

If we could work on things with more direct and immediate applicability and need, that might help clear up the funding problem a lot. There tends to be more money there.... Unless we can actually get an the wayb to a system that can actually produce fpgas and sell them for less than typical prices, and thereby make money of the fake if them directly.  Or the sale of the production equipment.

[Gerry]

I'm a proponent of "Open Source Everything", so yes, I'd be interested in a real project to do this. As I mentioned in a different thread, CBPP and in particular new financial institutions designed specifically for commons based enterprises.

Going directly to silicon fab does seem pretty ambitious. When I've looked around at Open Source Hardware, almost all of it needs to be targeted to FPLA CAD tools that are locked into particular commercial IC companies and the Open Source tools are still way behind. On the other hand, a little capital might go a long way in buying less than bleeding edge fab tech. Some of the CAD/CAE tools are getting there and we'd need to accelerate those development too (a Commons Based Bank maybe, I'm writing ideas on my "WAGN" http://gerry.wagn.org/ and I was just adding some text before checking out this group again).

I've always thought that people would have to be "paid" for this work, as Anthony has said elsewhere, we all have to eat. I agree that there is a lot of interest to do these things if we could figure out good ways to finance it. There is a lot of design and development work on the gory details and that requires that socially connected working groups are able to concentrate on the remaining problems with great intensity. You need to free the people doing this from concerns lower on Maslow's hierarchy of needs. By funding them to work in a commons we can accelerate the "design work for the future" and get more people, in particular people who like to tinker and build things better.

[John Griessen]

On 07/19/2018 05:08 PM, Gerry wrote:
> There is a lot of design and development work on the gory details and that requires that socially
> connected working groups are able to concentrate on the remaining problems with great intensity. > You need to free the people doing this from concerns lower on Maslow's hierarchy of needs.
> By funding them to work in a commons we can accelerate the "design work for the future" and get more
> people, in particular people who like to tinker and build things better.

My board design and enclosure molding project, https://github.com/kanzure/culture_shock,
has been helped out by volunteer programming
to go along with my electronics design, and some
patronage to help pay the bills while I crank on it, but still the hierarchy of needs comes back after
a while in the form of pressure to "pay more bills" and "go on vacation more often", so yes finance is a
big thing to any commons work. Else each person with the inclination to this
kind of work needs to have a jack of all trades set of skills and do short term work to pay for the long term goals.

Open source FPGAs would be a great thing, allowing designs to take into account all the well known timing
effects of any particular implementation instead of those being trade secrets. That plus a fab in an older
feature size by 8 year old equipment would be a boon to society.

Those are team efforts that are hard to get any where on without specialists. So financing needed to pay them
well enough.