Questions about zero-g chip fabrication and wafer size.

General

Distinguished
Apr 5, 2004
105
0
18,680
I read an article about solar power satellites. The article brought up the the topic about fabricating solar cells in orbit, which brings me to these questions.

1. Would zero-g and near-perfect vacuum help in chip fabrication?
Perhaps zero-g and free vacuum helps somehow.

2. What limits the wafer size?
Is it gravity?
Size of the equipment?
etc...
 

corvetteguy

Distinguished
Jan 15, 2006
1,545
0
19,780
I read an article about solar power satellites. The article brought up the the topic about fabricating solar cells in orbit, which brings me to these questions.

1. Would zero-g and near-perfect vacuum help in chip fabrication?
Perhaps zero-g and free vacuum helps somehow.

2. What limits the wafer size?
Is it gravity?
Size of the equipment?
etc...

The only thing i THINK i can answer is about wafer size, and I'd say cost limits that but i might be wrong.
 

MonocleCat

Distinguished
Apr 30, 2007
106
0
18,680
Have you considered how much it would cost to establish a manned CPU fab in orbit? I'm willing to bet that the costs significantly outweigh any benefit. It would be cool if they could build a completely automated fab up there that would simply build chips and shoot them down in canisters needing only to be resupplied every year or so. Maybe in a few years? <1nm space chips that run at 20ghz @ 30 degrees with 40 cores? Could probably recoup the initial investment if they sold these to enthusiasts at several hundred million each.
 

goldragon_70

Distinguished
Jan 13, 2007
731
0
18,980
#1 no dust and particles to get onto the wafers, more good wafers.

#2 No gravity, R&D would have to be implemented to make a design system that can work in 0G, since some substances need to be added.
 

General

Distinguished
Apr 5, 2004
105
0
18,680
Have you considered how much it would cost to establish a manned CPU fab in orbit? I'm willing to bet that the costs significantly outweigh any benefit. It would be cool if they could build a completely automated fab up there that would simply build chips and shoot them down in canisters needing only to be resupplied every year or so. Maybe in a few years? <1nm space chips that run at 20ghz @ 30 degrees with 40 cores? Could probably recoup the initial investment if they sold these to enthusiasts at several hundred million each.
The costs of a fab in orbit is not what this thread is about.
 

TeraMedia

Distinguished
Jan 26, 2006
904
1
18,990
The wafers are made by slicing up "ingots", which are big cylinders of silicon. It's probably difficult to make a bigger ingot with consistent purity and crystal alignment properties.

Ingots (and chips, for that matter) are made with the help of gravity, but you could probably simulate g's adequately by spinning a portion of the facility.

Since making an IC involves a lot of processes, and some of those processes could benefit from a vacuum environment, I bet those processes could be done better in space. Other processes might be impaired by the force of gravity, so being able to work in a nearly-zero-g environment would also help improve IC quality. It's certainly possible to simulate 1 g and 1 Atm in space; simulating 0 g and 0 Atm on earth is a lot more difficult, so perhaps that's where there might be a benefit to space-based fabs.

Until there's a space elevator though, the costs of shipping would far outweigh the marginal cost-benefit of IC quality improvement. Newegg couldn't really offer a UPS Ground option...
 

goldragon_70

Distinguished
Jan 13, 2007
731
0
18,980
Actually, You could probably hall enough chips down and raw supplies up in a shuttle that the profits would pay for the fuel, it's the running of the space station, and the fact that people would have to stay up there, plus the problem of Micro gravity, and the billions that would have to be spent on R&D that would kill the budget.
 

Mobius

Distinguished
Jul 8, 2002
380
0
18,780
Clearly you guys know a little about chip production and nothing about microgravity. There is no such thing as "Zero G" - it's a myth. Orbital satelites live in "microgravity" and not true zero gravity. It's complex - but trust me - it's true.

Dust is as prevalent in space as it is in your own home - but the ramifications of it are much more evident. It is a major problem on the ISS cleaning fungus and mould from things. (Fungus can eat glass, and even metals!) In fact recently when they removed a panel for inspection, they found a glob of disgusting brown water the size of a basketball floating in amongst the electronics!

So, the answer to the question is NO: there is no benefit to using lithography in orbit: the costs would be much higher, but more importantly, theres no process in the production of a semiconductor that will benefit from a microgravity production environment. In fact, if I remember my lithography 101, it might actually be harder, or impossible to use the current method in space.

Microgravity is ideal for growing chemically pure and perfectly formed crystals which can't be made on earth in exacting enough tolerances. What use this might be in the future is arguable. There's no doubt that there will arise product ranges which will benefit from being made in microgravity, but it is pretty pointless to speculate on what those might be.