Read more.Quote:
KINGPIN and TIN worked with EVGA on the project that has been in gestation for 7 years.
Printable View
Read more.Quote:
KINGPIN and TIN worked with EVGA on the project that has been in gestation for 7 years.
I know a few EVGA cards were catching fire, but this still seems a little extreme.
Not as extreme as the price will be :D
Let's hope this tinkering produces something worthwhile for the masses, not LN2, but some of the ideas of getting this to work could filter into something else? Anyway, to digress a little (sorry)... I have always wondered if there could be some kind of self contained gas chamber cooling system (liquid flows to CPU, evaporates into gas and also expels this heated gas into a chamber or even multiple chambers, where it then gets re-cooled and pumped back to the CPU as a liquid) that could be low maintenance like an AIO cooler but one that wouldn't also be a potential explosive weapon!
I believe you are referring to something called phase change cooling - there are platforms available to buy from LD Cooling for a couple of thousand Euro. Would post a link but think it may break forums rules - but its easy enough to google if you have a couple of grand sitting around to buy one!Quote:
Originally Posted by EvilCycle
So outside of marketing nonsense, how exactly is it supposed to work? Liquid nitrogen cooling as it's generally used in overclocking just works through evaporation - the liquid boils and the phase change takes heat with it. If you were to seal it in somewhere you prevent evaporation so just end up with some room temperature nitrogen sitting there (though that's really not happening here - N2 cannot be a liquid at room temperature and pressures for its supercritical phase at room temperature are immense). It's not some sort of physics-defying substance which just magically destroys heat!
So, it's basically just piping the nitrogen to the blocks rather than having to pour it in manually (probably, given the huge dewars either side of it) in which case it's not a closed loop?
Yep, heatpipes/vapour chambers are pretty standard in PC coolers. Liquid boils at the hot end, and condenses in the end attached to the fins.Quote:
Originally Posted by EvilCycle
Potentially you could pretend it's closed-loop, with a big chamber that slowly increases in pressure over a benchmark run. It's probably short enough for that to work, given the size of the two flasks nearby (with both getting pressurised, to make life easier for the pump). Given they aren't just trying to keep the chips at LN2's boiling point, so need to meter the flow precisely, automating the process is a worthwhile improvement (after all, who doesn't have automatic fan control on their PC?)Quote:
Originally Posted by watercooled
I'm not sure what you're describing?
Pressurised dewars like this hold only a fairly low pressure (compared to the triple point at least) in order to raise the boiling point accordingly and store/use it more efficiently, not to hold it as a liquid at room temperature (again, not possible). After it's been heated by the processors, you'd need something to remove the heat to re-condense it to make it a 'closed loop' system. Maybe they do re-cool it? You can't just do it using evaporation of more nitrogen though because then you gain nothing.
There's not much in the way of details in the video (plenty of random smart-looking diagrams and conveniently-located oscilloscopes doing nothing in particular) so it's hard to figure out what they're doing.
The thing is, you want the liquid to evaporate at the heat source but there are equally-sized pipes presumably carrying the fluid away from the blocks again, presumably as a gas here?
It seems like an automated liquid nitrogen cooling system, fine. But that doesn't make it closed-loop.
Google'ed for more information and came across an article at Gamer's Nexus which explains the process.
https://www.gamersnexus.net/news-pc/...ng-roboclocker
Quote:
One tank acts as intake LN2, the other acts as output LN2 (or gas).
So it's not closed and it's not even a loop!Quote:
... K|NGP|N is able to retrieve ~75% of his LN2 in the output tank, with the remainder lost to phase change
Remember this thing?
http://www.hexus.net/tech/news/cooli...change-cooler/
Now that's closed loop!
I don't care what it is, how it works or whether it works.
This is so unnecessarily overkill that I must have it.
All of your pwn shall belong to us.
Thanks guys, I seem to have a great talent in inventing things that already exist, maybe I should work for Apple?
I'll wait for the RGB version, thanks...Quote:
Originally Posted by philehidiot
Friend of mine keeps trying to design some kind of condenser-based cooler thing. Wonder if he knows about these...