Getting Crazy with Liquid HELIUM

Vector · Jan 11, 2001

OK, so I'm nuts and I don't have access to the suff yet, but maybe in about a year...anyway...

I bet if an entire sysem (motherboard, processor, ram, graphics, etc) was submerged in liquid He, you could get some phenomenal performance out of it...I mean we're talking like 4 degrees above absolute zero...

At that point most metals become superconductors, so the heat produced would be greatly reduced and electron flow would probably go up by about 10x that of normat...maybe more...

I'm thinking something along the lines of 500mhz fsb (1000DDR) and about a 5GHz processor...

The only problem that I could see at that temp would be microfractures in the silicone due to processor oscillation (brittle at that temp, you know)

Oh well, just a crazy thought...
Please feel free to reply with any comments

Rob Cork · Jan 11, 2001

The main technical problem with that sort of setup is getting the equipment down to that temperature gradually so it doesn't fracture. Just as a hot piece of glass can shatter if you place it in cold water, as the outer surface expands rapidly and the inner surface remains hot, a sudden drop from room temperature to 100's of C below freezing would put huge stresses on the components. I heard where some guys in Australia tried to cool it gradually by submersing their computer in Freon cooled by dry ice, and then once the temp had stabilised submersing it in liquid N2, but they still killed the mobo.

Plus you would have to constantly top up the liquid He unless you closed the system - but that would be dangerous because as the He was heated by the components it would easily boil, and the thing would explode. Maybe for a short time though, you could get some absurd performance out of it.

Nice idea, lots of practical problems to overcome - but it'd be damned impressive if it worked

Vector · Jan 11, 2001

I think the liquid nitrogen thing didn't work because nitrogen is slightly conductive in it's liquid state; however, liquid helium is completely non-conductive and really, really cold...

I'm a physics major right now, and we use liquid nitrogen to make liquid helium...grant you it's not the cheapest process in the world, but I may be able to persuade the department to give me some liquid He for my little experiment next year....

You're right about the stress fractures from instantly supercooling...and I did think about that, but I was concerned about the processor fracturing itself once it got down to temperature because it was oscillating so fast and it was so brittle...

Rob Cork · Jan 11, 2001

I see the concern with the processor fracturing - I have no idea whether that's likely, but it does seem possible. Perhaps a less extreme solution would be to use liquid helium/nitrogen as a secondary coolant, immersing the computer instead in some other liquid (an oil perhaps?) with a low freezing point (like below -150C - if such a liquid exists). You'd still have temps below -100C, though granted that's not quite superconducting range... yet ;-) That liquid N2 project I mentioned - it was only the mobo that they killed, everything else worked fine when back to normal temp, so maybe it was just the cold that killed the mobo, not the conductivity - who knows? Keep posting with ideas, it's cool to think about if not do it - I'm just starting a Natural Sciences degree (chemistry mainly), so this sort of stuff is pretty new to me but I'm learning all the time

.

Vector · Jan 11, 2001

Can't very well use a liquid that's not supercooling it...the superconductivity aspect of it is what makes it so appealing...otherwise you're just making it really cold but not cold enough to stimulate electron flow...

Anyway, a while back, I had access to a good deal of liquid N2 but I decided against trying anything because (if i'm not mistaken) liquid nitrogen is conductive...you would fry a motherboard in a second...liquid helium on the other hand would be perfect because it is completely **non-conductive**

I think that I would only have to get around the oscillation fracture idea...if that even turns out to be an issue

Big Mike · Jan 11, 2001

Actually the conjecture for why the board failed was that the electrolytic capacitors and or the battery ruptured when the liquid inside froze in the article i read, perhaps thats more the issue than anything else, the liquid N2 didnt seem to damage the CPU itself.

johnnyw · Oct 3, 2002

If you have the knowledge of how to do it but it still remains to get the liquid he, just plan your ideas calmly and draw graphs and all the necessary things just that when you get the materials, you can put down to work.
I think it is the first time something like this will be done.

FunkDaMonkMan · Oct 3, 2002

wow! ... if i were you I would try it.

If the Helium is cheap enough, you could test it on a kinda cheap system, and if it works then put a monster rig in there!

JigPu · Oct 3, 2002

Wouldn't causing superconductivity be kinda bad? I mean, sure it's great in some aspects (lower voltage, ect.), but the resistors shouldn't be superconducting at that temp unless they don't use carbon resistors, and so the resistance which hasn't changed would cause an incorrect voltage drop / current restriction because the resistance of the tiny traces has dropped from a few ohms (prolly more since they're so tiny!) to nil.

But even if the resistors do start superconducting (assuming the resistors are VERY thin pieces of wire), that would also be bad because then the resistance would be WAY too low. Might just blow a component or something....

Dunno if any of what I said actually holds true, but it's something that I just thought of.... I'm OK at electronics, but not that good

JigPu

FunkDaMonkMan · Oct 3, 2002

JigPu said:
Wouldn't causing superconductivity be kinda bad? I mean, sure it's great in some aspects (lower voltage, ect.), but the resistors shouldn't be superconducting at that temp unless they don't use carbon resistors, and so the resistance which hasn't changed would cause an incorrect voltage drop / current restriction because the resistance of the tiny traces has dropped from a few ohms (prolly more since they're so tiny!) to nil.

But even if the resistors do start superconducting (assuming the resistors are VERY thin pieces of wire), that would also be bad because then the resistance would be WAY too low. Might just blow a component or something....

Dunno if any of what I said actually holds true, but it's something that I just thought of.... I'm OK at electronics, but not that good

JigPu

Well resistors are made out of semi-conductors right?

do those become more conductive near absolute zero?

sry... i'm just 7weeks into my 1st chemistry class

Lusankya · Oct 3, 2002

Lets not forget.. no computer componet has been designed to operate at those temps.. so i doubt it would work

[EG]~NaTz~ · Oct 3, 2002

wow sounds outa my league submerging a comp i could see dropping some liquid helium on a heatsink hehe....

cwb27 · Oct 3, 2002

Can't remember if I'm thinking about the right thing here, but theoretically, doesn't all particle movement stop at absolute zero? (0K) or am I thinking of something different....

Lt. Max · Oct 3, 2002

cwd: but if absolute 0 is superconductor then i dont think particle movement could stop.. im just 16.. and i dont even kno half of what im talking about

-=HN=- Wild9 · Oct 4, 2002

cwb27 said:
Can't remember if I'm thinking about the right thing here, but theoretically, doesn't all particle movement stop at absolute zero? (0K) or am I thinking of something different....

the particle movement in the wire slows down so much that the electrons can flow more freely through the wire, that is why it becomes superconductive. If you take a magnet and dump some liquid nitrogen on it, then place another magnet on top of it, you can get it to just sit there and float. its pretty awesome to see. one year in my physics class my teacher got some liquid nitrogen for us to play with, thats some wicked bad stuff right there

but i could not even imagine some liquid helium. BTW isnt liquid helium around like 4K or something like that?

gravitywell · Oct 4, 2002

I think you should just buy a couple fans and pretend

Lusankya · Oct 4, 2002

cwb you are correct.. thats past the point that air turns into a solid on the floor.

Thing is.. we cannot acheive absolute 0 yet due to technological limits. Anything that gets to absolute 0 looses molecular cohesion and falls apart. We cannot get absolute zero because any container we try to get to absolute 0 would instantly fall apart.
On order to get to absolute 0 we would need a magnetic or electrical field to contain it in. Something thats not actually matter

TLKfatman · Oct 4, 2002

I think you should just buy a couple fans and pretend

lol i like your idea.

first of all no one hase ever gotten any thing to absolute zero gotten close

they say at absolute zero u could light a lite bilb and then jumper pos and neg together and take away the power sorce and it would stay lit for ever
here is the question tho if u were able to get a computer that cold would u need a power sup after start up?

glass · Oct 4, 2002

well. http://boojum.hut.fi/Low-Temp-Record.html
the record is 0.000 000 000 1 K.

liquid he isn't exactly cheap, and people already run into problems with the mobo getting too cold with ln2 cooling.

and one other thing, normal circuitry wouldn't probably work too well if one managed to get it to supraconductive because it would have to be designed with that in mind(taking into account resistances and delays and all that..).

edit: and the bulb wouldn't stay lit forever, since if it's emitting light it's using energy from the loop, though it wouldn't get lit if it was really supraconductive(the electricity would just circle in the loop forever).

Diggrr · Oct 4, 2002

Personally I'd love to see it happen and be proven wrong, but consider...If you change the conductances of components, and some become superconductors and some don't, then how would your computer keep the timings steady? Doesn't everything in the computer rely on the timing of each function? Would the current insulators on the components become condutors (though not super)?
Doesn't every material conduct at a different rate, thus the big search for the perfect material (again with the timing issue)?

And also, if there's no resistance in that lightbulb filament, then there's no heat/light produced. Besides, if you introduced that much heat into the "cold chamber" in that short period of time, the chamber itself would fracture/disintegrate, as well as not being able to maintain the temperature.
I follow the theory, but I'm a literalist most times

Getting Crazy with Liquid HELIUM

Member

Senior Member

Member

Senior Member

Member

Senior Head of Import Performance

Member

Member

Inactive Pokémon Moderator

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Underwater Senior Member

Similar threads