CPU At Absolute Zero?

[Benny11]

Sorry if this post is in the wrong section but I didn't know where else to post it.

I've always wondered what would happen if someone were to be able to cool a CPU to Absolute zero. I know (or have been taught) that semiconductors such as diodes and transistors become insulators at absolute zero and therefore would seize to work in a circuit. So if CPU's are made from millions of transistors... (They use silicone right?) then would a CPU seize to work at Absolute Zero as the cpu would now just be an insulator?

I also apologize if this is the stupidest question you have ever heard, it's just one of those things I have always wondered..

Thanks

[lawrencep93]

I don't think it would work, but it would be very hard to get it at absolute 0 and keep it there. Quite unpractical I'd say.

[3stars]

considering that nobody can get to absolute zero doubt it

[salochin]

I assume you'd get some sort of super-conduction happening, probably wouldn't be good

There's not really any point going that cold though - light can only travel so far in one clock cycle, and at 6 or 7 Ghz the distance it can travel is starting to approach the width of the processor. If you went much faster signals wouldn't make it from one side of the die to the other in one clock cycle.

[Concept CBF]

Originally Posted by salochin

I assume you'd get some sort of super-conduction happening, probably wouldn't be good

There's not really any point going that cold though - light can only travel so far in one clock cycle, and at 6 or 7 Ghz the distance it can travel is starting to approach the width of the processor. If you went much faster signals wouldn't make it from one side of the die to the other in one clock cycle.

Yep you'd definatly have super-conductivity, bascially making the processor the equivilant to a square of metal :P.

[fnp]

It's impossible to attain absolute zero.

[Little Man]

Not possible, but would be pretty cool though.
Thanks for derailing me off my homework

[karlcloudy]

Has anyone ever used Liquid Helium (I think thats what it is) as a cooling method? It's supposed to be colder than LN₂ I think, although I don't know a whole lot about science

[SLATYE]

Originally Posted by Concept CBF

Yep you'd definatly have super-conductivity, bascially making the processor the equivilant to a square of metal :P.

Superconductivity only occurs in certain elements. I don't think that silicon is one of them (regardless of how cold it gets). Copper, on the other hand, will superconduct at sufficiently low temperatures.

As others have said, semiconductors become insulators.

Rather more importantly, absolute zero implies that the system is completely disconnected from any other system (and the rest of the universe). The CPU has to be outside the universe if you want to achieve this, which will make using the computer an interesting challenge.

As Salochin said, even if the CPU did become 'perfect' (ie everything that was meant to conduct became a superconductor and everything else became a perfect insulator) you'd still hit speed-of-light limitations.

At 10GHz, light travels only 3cm per clock cycle. At this point the CPU may well be unable to actually remain synchronised (since many paths won't cross the chip by the shortest path).


karlcloudy: yes, liquid helium has been used. It's very expensive and difficult to use (because it's a superfluid and it tends to climb the walls of whatever container it's in) but it can be done.

[yanman]

There was a new milestone for "hot" superconductivity recently, I think it was something like 19 Kelvin?

[lawrencep93]

Originally Posted by salochin

I assume you'd get some sort of super-conduction happening, probably wouldn't be good

There's not really any point going that cold though - light can only travel so far in one clock cycle, and at 6 or 7 Ghz the distance it can travel is starting to approach the width of the processor. If you went much faster signals wouldn't make it from one side of the die to the other in one clock cycle.

I have seen a Pentium 4 overclocked to 8GHz and it still worked.

[Psychotik]

You cannot physically acheive absolute zero as there as is no way to remove every single piece of heat from an environment. I suspect that such extreme temperatures would render your CPU useless way before it ever gets to absolute zero (-273.15)

[SLATYE]

Originally Posted by lawrencep93

I have seen a Pentium 4 overclocked to 8GHz and it still worked.

For certain values of "work", anyway. It was running well enough that they could take a screenshot of CPU-Z. I suspect that if they used any reasonably complex instructions (that may actually require other parts of the CPU to be used) it would have failed.

Originally Posted by yanman

There was a new milestone for "hot" superconductivity recently, I think it was something like 19 Kelvin?

Much higher. They're past 100K. This means that liquid nitrogen (77K) can be used for cooling, rather than liquid helium.

EDIT: One team appears to have managed 260K, although that's still very experimental.

[Autti]

Most processor bug out well before absolute zero anyway, the AMD Tweaker is an except and that was successfully clocked to 6.4Ghz iirc with Liquid Helium.

[EvilGenius]

Originally Posted by Psychotik

You cannot physically acheive absolute zero as there as is no way to remove every single piece of heat from an environment. I suspect that such extreme temperatures would render your CPU useless way before it ever gets to absolute zero (-273.15)

It's not about removing the heat so much as it is slowing the molecular motion to a point where it can't impart any energy to anything else. That part alone would make running a cpu at absolute zero impossible.

( yes I know that removing the heat and slowing molecular motion are pretty much the same thing, so perhaps it would be better to say instead that the above is a better way to phrase it...)

If you somehow managed to achieve that state of molecular 'stillness', introducing an electrical current, basically turning the cpu on, would instantly disrupt that, unless of course as per slatye's post, someone dreams up a perfect insulator. Of course we're well into dream land by now