Slashdot Mirror
Doomsday PC-Cooling With Dual-Cascade Coolers
An anonymous reader submits "Four (4) compressors cooling one PC! Yes, it's big, yes it's heavy, yes it's loud and yes it does get your CPU and GPU cold, very cold. Is -100C cold enough for you? Cascade cooling is yet another chapter in a Finnish overclocker's neverending quest for optimal PC performance. Those things go down to -80 to -100C and can maintain the temperature. See here for the whole article with the pictures of the project."
But its the damn Gulf stream in the Atlantic that keeps (at least) southern Finland relatively warm, considering the geographic location...
The aim of science is not to open the door to infinite wisdom, but to set a limit to infinite error.
-Bertolt Brecht
Just that - Why?
Hey, I can appreciate water cooling. Keep the chip at basically room temperature, it increases its life and the OC'ers can push it a bit. But -100? WHY??? What possible use can this serve?
It doesn't even seem "cool" at this point (beyond the obvious pun). Wasting hundreds of watts, taking up way too much room (extra-large form-factor, anyone?), needing a fork-lift to move it... How does any of that benefit the PC or user?
Some things have an upper limit to what still constitutes "bigger/better/faster/harder". This definitely crosses that line with regard to chip cooling techniques.
Any cooling device creates more heat then it removes, due to some law ot theromdynamics.
Offset global warming? I think this thing might be a contributing factor.
"And The Geek Shall Inherit The Earth" --Jeff Darlington
In my book, 'optimal PC performance' involves consistant operation over a long period of time. This ignores the fact that the processor may be capable of more speed because of the nature of MHz ratings, but in the long run, this is 'optimal'.
This is interesting, and impressive, and admirable as an engineering exercise, but not exactly in persuit of mainstream 'optimal' performance characteristics. (Unless I RTFA and find that the processor does indeed last a good long time, and not burn bright and die out as I imagine it to under these conditions).
Maybe I'm just old fashioned.
"These people look deep within my soul and assign me a number based on the order in which I joined" --Homer re:
Lets face it, there are far more destructive ways to waste money than by attempting to make the fastest Home PC on the Planet.
Why climb the mountain? Because it is there.
Doing it just for the sake of having done it is enough, if that is what you want to do.
An infinite number of monkeys will eventually come up with the complete works of
While putting the computer outside is a stupid idea, I can think of a very simple system over there. Just drill two 1-inch wide holes in the wall next to your computer, attach two tubes to it and link them at the CPU fan level. This way the fan will get fresh air from outside to the CPU, and then throw it out. After all, that's what the compressor is all about: Getting something cold. The air outside should do it.
Of course, you need to drill holes in your house...
Write boring code, not shiny code!
I can't remember how low we went with the Transputers but they ran damn fast dunked in liquid helium. The processors did not reach the level of the helium because it was constantly evaporating.
The limit to overclocking is highly processor dependent. Some designs will simply end up in a race condition because some parts of the chip will work much faster than others and you end up missing the right edge of a pulse. Basically you give yourself a whole new region to discover timing errors in the design.
I don't think that the physical process is going to be a fixed limit, clearly this will be very dependent on the physical packaging. Chips are sent into space to face some pretty unpleasant temperature ranges.
Depending on your material there is a point when your band gap goes all wonky and things start breaking down. Most times what you are worried about is the effect in the high temperature region, but there are equally wierd things in the low temperature region.
This is definitely not something that is recommended for most applications. There are a couple of oddball ones, like cryptanalysis where it is really hard to get a result but once you get one it is trivial to check. I would not be surprised if GCHQ has a swimingpool sized machine for brute force key cracking dunked in some type of cooling liquid. The NSA would just chuck money at the problem.
Looking for an Information Security student project suggestion?
Try http://dotcrimeManifesto.com/
For all those wailing "Like, WTF?" and "This isn't worth it!" I'll say this once:
Well duh. Do you think we don't understand the value of time, space, and money and can't do an investment/return calculation?
This is cool because they can do it. It's on Slashdot because lots of us think it's nifty to turn a 2.2GHz processor into a 4+GHz processor.
Yes, it does cost more, take more space, and more time to set up than two 3GHz machines, or even a dual processor 3GHz.
But it's like my high school instructor telling me 10 years ago that making a microcontroller controlled light dimmer is non-trivial. I did it then, and it requires fewer than 25 lines of assembly code on a simple microcontoller. Was I geeked when I finished? You bet.
People are constantly trying to break records, and this is no exception. The higher the record is set, the more effort and resources must be put in to beat it.
-Adam
Why is it becoming harder to post on slashdot? 4/5 of the time I get an incomplete page when I press submit or preview.
Extreme overclocking often takes the fastest available processor and then overclocks it as much as possible. The goal is not the most bang for your buck, but purely the most bang.
What money? He said he got both cryofreezers for free, because they weren't workign originally. So the only cost is his time.
Not quite, yourself.
Consider a heatsink. It removes lots of heat from a CPU, but generates no heat at all. That said, any devices which cools something beyond ambient temperature will generate heat of its own, which is nearly what you said.
High-speed Road Trip (18.000KPH)
The first thing I thought is this is too complicated. A single LN2 compressor, some insulating tubes running into the box, and a heat exchange instead of a heat sink, you could easily chill that baby to 150K. The compressor would not even have to be in the same room. You would even have to charge it often if you kept the N2 clean.
"She's a scientist and a lesbian. She's not going to let it slide." Orphan Black
Extreme overclocking often takes the fastest available processor and then overclocks it as much as possible. The goal is not the most bang for your buck, but purely the most bang.
Yes, but as I said in my original post, wouldn't the "most bang" come from buying other CPUs, instead of making the one(s) you have run X% faster, with that money?
I'll bet that cooling rig is worth a bunch of Opterons...
Treehugger? Treehugger... Treehugger!
I'd link you to a site explaining humor, but I think it would be a lost cause.
I browse Slashdot at +3, Funny
If you do this, be sure that you water-harden your motherboard around your processor.
Nasty airborne water molecules condensing on my hardware...snort.
Well, he got the equipment for free (give or take fixing it: RTFA). As it stands, not a very practical general solution, but a really cool (pun intended) one off that may spur some new research somewhere.