And you thought 3Ghz was fast...

[kram 2.0]

http://slashdot.org/article.pl?sid=04/09/26/0019229

My computer feels slow...

kram

[Yuanji]

holy jeebus! dang..... now what happens if the liquid nitrogen tube decides to "break" one of these days...
Edit-i wish they broke 6ghz with an athlon 64 now tht would be fast!

[RenamonXP]

thats so twisted its sickening...

[dontstay]

Originally Posted by Yuanji holy jeebus! dang..... now what happens if the liquid nitrogen tube decides to "break" one of these days... Edit-i wish they broke 6ghz with an athlon 64 now tht would be fast!

that would be really fasttttttttttttttttt zoom...............

[Tin]

LN2 suicide run.....stop giving me ideas Kram .

[dontstay]

25sec superpi

[kram 2.0]

Originally Posted by Tin Canary LN2 suicide run.....stop giving me ideas Kram .

lol - sorry about that Tin - won't be surprised to see you actually do it though . But man - 6Ghz...with the money it took to get all that equipment, I would've gotten myself a ATI Radeon X800XT, better yet a regular Intel P4 3.6E and have kept it that way . Man - imagine the computing power.

kram

[Yuanji]

imagine the amount of power required to run that cpu on all day

[roomwithamoose]

I want to know what kind of temps they had that thing at! With LN2, you could get temps well below -300F. I want to know how a cpu and mobo can hadle being that cold. And what did they do about all the moisture?? Man, this has got me REALLY curious. Anybody out here want to team up and try it with a 64?

[kram 2.0]

Originally Posted by roomwithamoose I want to know what kind of temps they had that thing at! With LN2, you could get temps well below -300F. I want to know how a cpu and mobo can hadle being that cold. And what did they do about all the moisture?? Man, this has got me REALLY curious. Anybody out here want to team up and try it with a 64?

Don't know if the Athlon 64, which runs hotter than the Pentium 4 at correponding clock speed (ie. compare Intel P4 2.0 to AMD Athlon 64 3400+ I think), can do that. Even with that, it runs at .13 nm process - the .09nm process is what got the Prescotts to overclock that fast...

kram

[timbit]

90 nano a64's are on route.... be interesting to see what they can do.

[Punked Out Comp]

WOW thats sweeeeeeeet! Are there any parts that wouldnt bottleneck that much power?

[roomwithamoose]

Originally Posted by punked out comp WOW thats sweeeeeeeet! Are there any parts that wouldnt bottleneck that much power?

Seriously, what kind of RAM were they using?

[discodanman45]

Probably overclocking some DDR2 PC2-5400. Well I am sure that in a few years we will all have cpu's at 6 GHz and there will be some psycho who needs a 10 GHz cpu.

[imbest123]

I remember the cpu to ram ratio was 3:2 but even at that its still impressive. New record holder lol.

[Punked Out Comp]

I'm going to do that one day. A few questions though.

Aren't there limits on how far you can overclock a cpu?

No matter what kind, liquid conducts electricity, so wouldnt the LN2 short circuit the cpu?

Can you fill a whole watercooling tank with LN2?

What about condensation?

[Punked Out Comp]

*Bump*-I'm really thinking about doing this with an fx53 so Can someone please answer my questions.

[roomwithamoose]

An LN2 cooling system is completely different from a water cooling. You should try and contact the guys who did it, see if they can give you any tips/hints. I would do it right now, but my money needs to go other palces first. I rig like that would cost a pretty penny, not to mention how many things you'll break in the process of perfecting it. If you've got the cash, just do it and don't be afraid to break stuff.

I'm pretty sure that the LN2 is never touching the cpu directly, just like a water cooler doesn't. THere has to be some sort of LN2 black that straddles the cpu. And don't forget that the cpu isn't the only thing that needs cooling, you've got all the other components on the mobo to worry about.

Condensation is what I think the hardest would be. How do you get rid of it? There will be signifigant condensation no matter what kind of insulation you have.

And as to a limit for ocing a cpu, there certainly is. The temps are what would limit the cpu, so if you can keep it cool, then you're set. Not only that, but the colder a conductor gets the more effiecient it is. I wonder how many cpus and other components they burned.

[kram 2.0]

Careful - as the AMD Athlon 64 FX series currently are at 13 nm process, you won't be able to go as high as the Intel P4 Prescotts have...Intel has the 65nm processor already planned out - I'll need to see how that will turn out though - it'd be interesting to see how a 64-bit processor will handle the OC'ing, though.

kram

[mattg2k4]

>Aren't there limits on how far you can overclock a cpu?
OC'ing has its limits. Most don't have to worry about even coming close.

>No matter what kind, liquid conducts electricity, so wouldnt the LN2 short circuit the Not all liquids conduct, oil for example is used as an insulator. I don't know about LN specifically but if I had to guess I'd say it's an insulator because of it just being two N atoms w/ a single bond.cpu?

Can you fill a whole watercooling tank with LN2?
You can do whatever you like if you have enough LN.

What about condensation?
Assuming I'm right and LN is indeed an insulator, then by surrounding the system completely with the LN no water can get in, thus no condensation. Barring that I would use something that is applied as a liquid and then solidifies. I seem to recall various OCers using nail polish of some sort.

[Punked Out Comp]

If its an insulator does that mean if you made sure somehow that there wasnt a *single* crrack, you could just fill the case up with ln2 without covering the components?

[timbit]

i'm thinking that it's a little more complex than just filling a case with it.... seems to me like the case would have to be totally evacuated of any air (because air contains water, and you wouldn't really want ice forming in there...) and then filled with the stuff. complex procedure. not only that, you would need some way to ensure that the liquid didn't get in between components (ie somehow get in between cpu pins and socket, etc....) because if it is an insulator, then it could interupt flow of electricity.... and this is all assuming that its an insulator. not only that, i sincerely doubt that all computer parts are rated for temps that low. making everything that cold could cause some parts to become very brittle, and any jarring could actually destroy circuits (i believe). this is all conjecture of course, but it makes sense to me.

[roomwithamoose]

Here's an interesting link. Geez, it says that they were running an Althon to -200F. That's pretty cold.

Regardless of LN2 being an insulator, it is very difficult for any of us to create a computer that is a virtual vaccuum. Being that, condensation will definitely be something to worry about. And also the point that timbit came up with shows that you can't just pour LN2 on your mobo. The first reason is that it will obviously halt any current, and I assuming that most of these components could not handle -300F. As seen in the above link, they did you a block type system.

[Liquidparadox]

Here is my 2 cents. I would think that there would probably be very little condensation between the chip and the liquid nitrogen. You have to remember that the only way that you can keep nitrogen in a liquid form is to keep it at very high pressures, otherwise it just ends up subliming into just a cloud. Heck I think that the boiling point of liquid nitrogen is something like -200 degrees F. Realistically, Nitrogen does not occure in nature as a liquid, it is only a solid and a gas, and it goes from a solid to a gas by way of sublimation. By that way of thinking, condensation should never occure since it is not a natural cycle of nitrogen.

[roomwithamoose]

You don't have to worry about the LN2 condensing, you have to worry about the water in the air surrounding the container. It's just like a glass of iced tea, the tea is colder than the air outside, so when the water in the air is cooled below dew point it condenses onto the glass, making it wet. To prevent this, I would think that you'd have to make sure that no air was touching the container of LN2, which they did with foam. But I wonder if you'd have to find some way to stop condensation on the cpu and socket, since it's being cooled so much.

[Tin]

using LN2 is a suicide run, as i stated above. LN2 evaporates extremely fast and most processors burn out in a matter of minutes. most LN2 runs only last long enough to run a few benchmarks and get a few screen shots before all of the LN2 is evaporated away. LN2 is a far from practical cooling solution. think of it this way, would you drive an alcohol burning top-fuel dragster to work everyday? its pretty much the same idea, youll only get so far before somehting breaks. a much better solution would be to get a Vapo-chill or a Prometia, but those cost as much as a complete system for the most part. another thing you could do is get a hot plate (thermo-elecric cooling solution.....aka peltier) and mount the hot plate in between your waterblock and CPU. the way a hot plate works, the side resting on the CPU is cold, and the side the waterblock rests on is hot. you need an extremely powerful PSU to run a peltier, since a peltier itself can consume 200W or more.

[roomwithamoose]

That's a very good point Tin. But theoreticaly, if you were able to create a LN2 cooling system that was completely compressed, wouldn't that take away the chance for the LN2 to evaporate? Albeit, that would be an extremely costly cooling system, but it seems like that would work to me.

[Tin]

Originally Posted by roomwithamoose That's a very good point Tin. But theoreticaly, if you were able to create a LN2 cooling system that was completely compressed, wouldn't that take away the chance for the LN2 to evaporate? Albeit, that would be an extremely costly cooling system, but it seems like that would work to me.

the only problem is that you need a license to purchase LN2, because its such a hazardous material.

[Yuanji]

realistically, would any of you people be willing to pay that much for a computer that can process that fast? would that much processing speed actually be useful to the home consumer???

[roomwithamoose]

Well, actually the first problem (for me, at least) would be coming up with the money for the cpu!!! But after that, tons more hurdles. And even if I did have this kind of money, I wouldn't do it unless I needed that kind of processing power. And I'm not about to have a need to calculate the velocities of different solar wind flows between multiple solar systems anytime soon, so I won't worry about it yet. It's just fun to talk about! Now the Vapochill systems, THAT is something I'd do!