Project Create microprocessor using household equipment?

[Beholder]

In the event that Taiwan's chip foundry is destroyed to prevent the PRC from getting a technological monopoly, the price for microchips would skyrocket until most of us can't afford them. There are ways to create basic circuitboards by etching from laser printers, resin printing a stencil, or simply 3D printing wires directly with conductive filaments. These would however not have the density required for a processor. Another option would be to reprogram a Bluray burner into an etching machine, but creating functional transistors requires multiple layers with perfect overlap, and finding a material that can work as both source, gate and drain using a laser induced chemical reaction. Once that step is complete, you would obviously try to keep things simple without hyperthreading nor SIMD extensions, like an ARMv6 core without VFP.

Do you think that it's possible to create your own processor at home using today's consumer electronics as the tools, or is it still just a dream to create everything that we take for granted during a chrisis?

 

[Einstein]

If you had lots of money I suppose you could make an integrated circuit. But making a processor chip might take a real big wallet. Because of the density of components I would bet you would need a billion dollars to do it.

 

[Wind7]

If you had lots of money I suppose you could make an integrated circuit. But making a processor chip might take a real big wallet. Because of the density of components I would bet you would need a billion dollars to do it.

Not if you had a couple of coconut shells and some bamboo.

...If The Professor can do it, So Can You!


"Then again....You can make a clothes washer......Fashion bakeware out of c-4 explosive
and re-train a robot to send for help.......BUT.....You Can't Fix A Hole??!"

 

[MODAT7]

I do home PCB etching for some of my electronics projects involving microcontrollers. It's extremely cranky for thin traces. Nearly all blank copper clad PCB's being made in china doesn't help either.

A silicon furnace to make integrated circuits, the simple ones for ASIC prototyping and testing, cost as much as a house and only get more expensive. If you tried to emulate the evaporative deposition process with a laser, you'd at least need a clean room, a vacuum chamber, some very pure and expensive chemicals that would make the deposition (that are also probably toxic and biohazards), some type of high resolution deposition system for the chemicals that would put the correct amount down, and a super accurate positioning system for the laser head. I doubt it could be done with a laser as the intense heat needed in the chemical reactions would partially melt the layer below it, and the highly localized thermal expansion would crack the connections around it. If you think in terms of a molten metal inkjet printer, that might be more viable, but still has some thermal issues.

It sounds like a neat and revolutionary project if someone could figure out how to pull it off, but I don't see that happening within the next half century... at least with technology that's "out in the open".

For FPGA's, there are relatively simple and open source processor cores by the RISC-V and OpenRISC projects. Some of these variations have been taken into ASIC silicon with success. I'm guessing there are tens of thousands of gates involved, and each gate would be at least 3 transistors, maybe even 5. Even if you could pull off some home chip furnacing at mid 1960's levels, the finished processor would probably be as large as a desk and take a kilowatt to operate at the speed of a fast calculator.

For those concerned about chip shortages out of Taiwan and China in the event of a war, the only real thing to do is stock up on parts you think you'll need before TSHTF. It's also good to learn how to salvage parts out of old electronics and computers. Even dead ones aren't really 100% dead and have a lot of expensive parts that can be recycled into new projects... which I often do. It's usually not too hard to look up data sheets if part numbers are clear.

A more simple and cost effective solution would be to take a page from Putin's playbook in Ukraine using Iranian drones to "cause problems". Program a high speed, long distance drone to "take care of coming governmental problems before they become a problem". I've had some ideas about drones for awhile and have kept quite, but now that they're out in the open, they aren't considered very revolutionary anymore.

 

[Beholder]

I do home PCB etching for some of my electronics projects involving microcontrollers. It's extremely cranky for thin traces. Nearly all blank copper clad PCB's being made in china doesn't help either.

A silicon furnace to make integrated circuits, the simple ones for ASIC prototyping and testing, cost as much as a house and only get more expensive. If you tried to emulate the evaporative deposition process with a laser, you'd at least need a clean room, a vacuum chamber, some very pure and expensive chemicals that would make the deposition (that are also probably toxic and biohazards), some type of high resolution deposition system for the chemicals that would put the correct amount down, and a super accurate positioning system for the laser head. I doubt it could be done with a laser as the intense heat needed in the chemical reactions would partially melt the layer below it, and the highly localized thermal expansion would crack the connections around it. If you think in terms of a molten metal inkjet printer, that might be more viable, but still has some thermal issues.

It sounds like a neat and revolutionary project if someone could figure out how to pull it off, but I don't see that happening within the next half century... at least with technology that's "out in the open".

For FPGA's, there are relatively simple and open source processor cores by the RISC-V and OpenRISC projects. Some of these variations have been taken into ASIC silicon with success. I'm guessing there are tens of thousands of gates involved, and each gate would be at least 3 transistors, maybe even 5. Even if you could pull off some home chip furnacing at mid 1960's levels, the finished processor would probably be as large as a desk and take a kilowatt to operate at the speed of a fast calculator.

For those concerned about chip shortages out of Taiwan and China in the event of a war, the only real thing to do is stock up on parts you think you'll need before TSHTF. It's also good to learn how to salvage parts out of old electronics and computers. Even dead ones aren't really 100% dead and have a lot of expensive parts that can be recycled into new projects... which I often do. It's usually not too hard to look up data sheets if part numbers are clear.

A more simple and cost effective solution would be to take a page from Putin's playbook in Ukraine using Iranian drones to "cause problems". Program a high speed, long distance drone to "take care of coming governmental problems before they become a problem". I've had some ideas about drones for awhile and have kept quite, but now that they're out in the open, they aren't considered very revolutionary anymore.

Eagerly awaiting the day when we can buy a hardware license online and 3D print the latest graphics card from trash. But maybe self replicating computers with DNA and human brain cells would come first.