out just of pure curiosity how tiny can a computer chip maker can get? like to make a single z80 CPU or something

back in the heydays of 3d printers, they were huge big things but now you can buy one at best buy for a week worth of pay.

6492585
Let's see...
Zilog designs and manufactures the Z80 series and its derivatives
AMD just designs the chips and sends them to Globalfoundries to manufacture. Globalfoundries used to be part of AMD, but was spun off as its own company.
Intel designs and manufactures their chips
Samsung designs and manufactures their chips
Apple designs chips for Samsung to manufacture, though I've heard that they switched to I think TSMC because of patent disputes.
TSMC manufactures chips for a variety of companies.
Texas Instruments, IBM, National Semiconductor, Atmel, Fairchild Semiconductor, etc. all design and manufacture their own chips as far as I know, but I remember that some of them also use outside fabs such as TSMC to take off some of the load.

And then there's Jeri Ellsworth and her homemade chip lab.

That's all I can think of at the moment.

6492585
Wait, ARM. They used to be a british chip designing firm, but now they're part of the Japanese Softbank conglomerate of companies.

6492617
i was more thinking of is it possible to just put silicone and other things that make a chip into the thing and you wait a good time before it put out one chip

6492627
Silicon, not silicone. One is a semiconducting metalloid, the other is a soft rubber used for breast implants.

And no, if it were that simple to make integrated circuit chips, then companies like TSMC would go bankrupt almost immediately, given that it would lead to all the chip designers simply manufacturing chips themselves.

Also,

just put silicone and other things that make a chip into the thing and you wait a good time before it put out one chip

Put silicon and other things that make a chip into the thing? What do you mean?

6492642
i know to make a computer chip you need more the silicon you need rare metals like copper, gold and other to make the logic gates

6492715
No, copper and gold are for bonding the chip to the packaging, not the actual chip itself.

Here are some of the elements that are used in the actual chip fabrication process, though silicon by itself is still usable for simpler circuits.

6492715
Why are you suddenly so interested in this field, anyways?

6492730
pure curiosity like i say plus i like to make things on the side for the fun of it.

6492743
Took Jeri several years and quite a few scrapped and hand-built parts to get even the simplest circuit going.

6492585
Big.

I once saw one about the size of a 3D printer, so there's that.

6495656
Integrated circuit manufacturing technology requires multiple series of crucibles, ovens, engraving/doping machines, etc. I highly doubt that all that could fit in the space of a 3D printer.

6495820
I'm sure that people thought that about some of the things that some types of 3D printing/additive manufacturing can do as well.

6497220
No seriously.

This is one manufacturing facility. It's not gonna be able to physically shrink to 3D printer sizes anytime soon, unless you want extremely low-quality chips. Integrated circuits must be built in clean rooms, or else even a single speck of dust can destroy the entire process.

Don't get me started on how many different materials outside of silicon are used as well.

3D printing was only able to be miniaturized because it uses a single type of material to create a low-detail model with many millimeters worth of impurities. Integrated circuits are manufactured using hundreds of different compounds, on a nanometer scale nearing single-atom sizes.

6497220

More fab labs, or manufacturing facilities.

6497220
If you were to theoretically create a 3D printer sized machine that could fabricate integrated circuits, (forget CPUs and GPUs, at the current level of technology a homemade model would hardly compete), then how would it be constructed, and how would it function?

6497226
6497228
6497229
I'm not arguing that the places don't exist, or that they can't be used to get the job done; i'm arguing that they are not the only things that can get the job done; like, if i showed you a image of a metal refinery you wouldn't stop believing that there are metal 3D printers would you?

6498535
Once again, 3D printers and chip fabs are two entirely different things. I know that both metal 3D printers and metal refineries exist, both for the purpose of metalworking, but once again, it's currently physically impossible to shrink a chip fab into anything smaller than at least a basement's worth of machinery.

You didn't answer my question, either. Give me evidence of a 3D printer-like machine that could manufacture integrated circuits. Designs, scientific papers, patents, etc. are fine.

6498551
...Bitch i just gave you six; tell be one part of a PCB that requires such a large machine to fabricate.

6499449
Oh, you're talking about PCBs, not integrated circuits. My bad. The original poster was talking about integrated circuits such as the Zilog Z80 or the Intel Core i-series. Printed circuit board can indeed be made using 3D printer-sized machines due to their relatively low complexity.

Sorry for misunderstanding. ^_^

6499449
For reference, these are integrated circuits.

The last image shows the opened package of a 74181 ALU chip, which was one of the precursors to modern-day CPUs. As you can see, the logic is much more densely packed than a PCB.

6499460
6499462
Oh, yeah, i see that now; still of the oppinion that you could totally make one of those things about the size of a printer using some high high sensitivity parts and some vacume chamber equipment, since even electron microscope equipment isn't as large as what you're describing and they can actually move individual atoms around into micro structures, and the result would probably work rather slowly (to the point of taking days to do what the mass production would take minutes for), but i can totally see where you're comming from with it not yet being a thing.

6501048
Yes, in aware that it's possible to push individual atoms into place using things like that.

What I was more specifically speaking about was integrated circuit production from a mass-production standpoint. Basically, pushing single atoms in that way wouldn't possibly be economically feasible.

Once again, sorry for any misunderstanding, but i can see your point now. Let's just end this conversation here, alright? ^_^

6501234
I've been having a good time with this conversation; i don't see any reason to end it if you don't want to, and yeah i know that individual atom restructuring would be massively inefficient; i was thinking more along the lines of how, if that level of precision was possible, then it would stand to reason that a less precise one with a material printing mechanism attached to it should totally be viable. Maybe using some form of metalloprotein that react to light (which would then exploit laser tech to form highly detaled structures) by turning into something that is both conductive and doesn't react to a enzyme that could remove unreacted substance?

I'm mostly just spitballing here at this point; feel free to just ignore me and i'll likely forget about this thread.

6502083
I've seen experiments like that from IBM. They once made a short video using nothing but an electron microscope and individual atoms. So, as long as you aren't planning on mass-production, then yes. It's indeed possible to manufacture something as complex as integrated circuit chips using a 3D printer-sized machine.

6502097
Might want to give it a wide brush attachment/setting before you try that, and maybe switch back to something narrower for some subtractive manufacturing to clean it up. Seems like it would be the best way to go about it for time efficiency and product effectiveness.

Anyway, on the question of the actual size of it...I think that electron microscopes are about the size of a smallish car? Like, would-fit-in-a-van-if-you-gutted-everything-but-the-drivers-seat-first sized? Though i'm not sure how much of that was seals and i'm probably remembering wrong.

6503608
Yeah, they're roughly the size of a small car. I've seen a couple while I was touring the University of Utah. As for manufacturing the chips, yes, it could potentially be possible to make one using a 3D-printer like machine if you revert to a much larger process size. That way, if the transistors are measured in millimeters rather than nanometers, a very precise 3D printer may be able to print down circuits on a chip.

6503620
It's probably possible to cut the size down to like half of what it is if you don't mind working with precision on the level of groups of buckyballs instead of individual atoms. Do you think that it would be easier to print down largish "ribbons" of conductive material and then carve them into individual...Whatever they call the connections between logic gates, then it would be to print down individual ones?

I really like the idea of personally owned fabricators in case you haven't noticed. Especially self replicable fabricators. Just the idea that anyone wanting to make a profit out of renting out fabrication runtimes (? I think that's the right word?) would need to compete with the cost of just getting your own one; it makes me want to be around to see the face on old money type people when they realise that no one is using their currency any more~

6504459
It's possible to build your own. Remember, I linked to an article where a person by the name of Jeri Ellsworth was able to scrape together the correct machinery to produce crude circuits.

And if you want to go by hand, here's the first ever integrated circuit for reference.

It was made from a hand-cut wafer of Germanium, which while isn't as high-quality as Silicon, is able to withstand many more impurities, making it a practical choice if you want to make homemade circuits. In a way, this was constructed in the exact way you described. The chip was cut into a strip off of a larger wafer of material, and then carved and soldered together into a working device. While it's definitely not as efficient, cost or material-wise, it's the simplest possible way to create a circuit that has any chance of working at all.