I like the sounds of this! if it works on PSUs GFX etc then it should take the industry by storm
if the ‘liquid metal’ that they're using is mercury, they are going to fall foul of RoHS and WEEE, which will make it very difficult to be able to sell it, not to mention the fact that if it were to leak you'd get mercury poisoning
Could potentially be rubidium, francium, caesium or gallium.
rubidium, francium and ceasium are too reactive, and explode upon contact with water, ruling those out, and if gallium is used, if it gets too cold it expands when it freezes and this would break the container
I suspect its none of the above due to the extreemly hazardous nature or high cost of that. My guess is that it would be some form of compound suspension ( like a ferrofluid ) , but thats off the top of my head.
Where is it transfering the heat to? From my understanding the “liquid metal” is just going around in circles keeping the heat within the system?
Its being released as it passes the fins.
its being used in the same way as a heatpipe, but its just much better at moving heat than a heatpipe. so its moving heat into the metal fins and then a fan will probably be attached to move the heat away
What about cost of this very interesting yet new technology.
Will it be cheaper than a regular basic water set up than you can buy??? Will it still require a fan to push heat away!
Just thought I should pop in to emphasize "the worlds first
commercially available liquid metal based CPU cooler". Liquid metal cooling for CPUs is not a new technology, this is just the first time it has been offered to the commercial home user market. Liquid metal cooling has been available for mission-critical server applications for years, using a mix of gallium & some other stuff (indium I think?) pumped around the loop with a magnetohydrodynamic thruster. The one I remember seeing on eBay a couple years back for Xeons resembled a normal water cooling loop more closely, with a separate radiator/fan, pump, CPU block & tubing, rather than an all-in-one tower unit like this one.
There have also been previous promises of consumer-grade liquid metal cooling before, such as Sapphire's
Radeon Blizzard X850XTPE.
Brewster0101
Will it still require a fan to push heat away!
I can't imagine it could operate without a fan. If it draws heat from the CPU more efficiently, it's just going to dump it to the fins faster, so it'll still need transferring away from the fins.
it'll need a fan, the fins are too dense to work in passive very well, and you can clearly see its designed to fit fans on either side.
hmm does sound interesting, need to see some figures soon though, ie price n temps.
Moby-Dick
http://en.wikipedia.org/wiki/Galinstan
et voila :)
used instead of mercury in modern thermometers,
“It is also a promising coolant, though its cost and aggressivity are major obstacles for its use.”
a low quality mercury thermometer costs about £10, and they contain only about 1 cubic centimetre of mercury or less, and guessing how much the pipes on this cooler will contain, this is probably going to cost at least £100, maybe double or triple that (galinstan may be more expensive or cheaper than mercury, but I'm assuming there isnt much difference). I think that its going to be contending with mid to high end water cooling in both price and performance.
wait until someone cuts one open, it may just be a sales ploy and a really thin pipe that doesnt do much at all apart from remove money from the bank balance :)
GoNz0
wait until someone cuts one open, it may just be a sales ploy and a really thin pipe that doesnt do much at all apart from remove money from the bank balance :)
I doubt they'd risk being sued for false advertisement. It's quite easy to see how it works also, as the liquid metal reaches a higher temperature, it's viscosity increases and the electromagnetic pump pulls the warmer metal up easier which in turn pushes the cooler metal down.
Not sure why a fan wouldn't be needed, the fanless, no moving parts pump simply moves the heat around a normal sized/shape heatsink, you still need to move heat from the heatsink to the air and then out of the case. The equillibrium point between this and any other cooler won't be much different using the same temperature air and a chip putting out the same amount of heat. THe only way to improve cooling is to reduce the overall temp of the heatsink or at least balance out the temps as the base and lower heatsinks area's should be hotter than the furthest parts of the heatsink, if they are the same temperature there is no room for improvement.
Need someone to do a review of a standard big tower sink and check the difference in temps between the parts closest to the base, and the parts furthest away, that will show us how good the spreading out of heat is being done.
I do wonder what happened to those TMD fans as theres still room for improvement for higher airflow quieter fans, less failure and a much smaller deadzone due to the middle of the fan being smaller on the TMD's. Oh well.
Biscuit
I like the sounds of this! if it works on PSUs GFX etc then it should take the industry by storm
Yeah it's an interesting piece of technology but I have doubts it will ever become mainstream. the biggest reason was mentioned by lost eden, the Sapphire Blizzard card - a company did all the work required to make a functioning prototype with the promise of mass production, but dropped it, presumabley because something didn't actually add up, at least for graphics coolers.
shaithis
Could potentially be rubidium, francium, caesium or gallium.
Possible, but highly unlikely it's any of those, at least in their pure state, simply because they all melt above 25 C, and room temperature is usually taken to be around 20 C. It would mean that your coolant would first have to melt before your cooler can work properly, and since the thermal conductivity of all these metals are 10+ times worse than things like copper your chip will be reaching the 80 C range before the cooler has even started flowing properly, just so there's enough of a temperature difference to transmit the melting energy required…
hmmmmm
If anybody finds a specific heat capacity figure for Galinstan post it back here and I knock out a rough calculation of operating conditions along with a comparision to water cooling and the like (I'd look now myself but food calls :P).
There's no need for a fan thanks to the electromagnetic pump, and unlike water cooling, there's no need to mount radiators, pumps or other equipment. With no moving parts, noise isn't a factor, either.
And how's the heat supposed to leave the fins? As far as I understand the principle, the electromagnetic pump simply circulates the liquid metal eutectikum in the ‘heatpipes’. Why else would that thing have such a staggering amount of (extremely) closely spaced fins?
Cistron
And how's the heat supposed to leave the fins? As far as I understand the principle, the electromagnetic pump simply circulates the liquid metal eutectikum in the ‘heatpipes’. Why else would that thing have such a staggering amount of (extremely) closely spaced fins?
I was thinking along those lines. I'm sure you will need a half decent air fow to get it to work very effectively though I am interested to see how effective it is :)
If it does work it will be interesting to see how compact they can make the technology, for HTPC's etc..
There are alloys that melt at lower temperatures than gallium, they usually contain gallium them selfs though. They could be easily used for this application.
Edit: Such alloys typically consist of gallium indium and tin.
it's a good idea, and if it works would be great. The only problem is is going to be how much it costs.
DataMatrix
I doubt they'd risk being sued for false advertisement. It's quite easy to see how it works also, as the liquid metal reaches a higher temperature, it's viscosity increases and the electromagnetic pump pulls the warmer metal up easier which in turn pushes the cooler metal down.
no, the temperature shouldn't have an effect on the performance of the motor, if it was an open system it might, but as the pipes are sealed any change in temperature will affect the pressure but not the density of the liquid.
the electromagnetic pump just works like an ordinary pump and just circulates the liquid around the circuit continuously, the only difference is that it is using electromagnetic induction -which relies on the liquid conducting electricity- to push the liquid
shaithis
Could potentially be rubidium, francium, caesium or gallium.
The most stable isotope of Francium has a half life measured in minutes and is a solid at room temperature..
Caesium explodes in when it comes into contact with water and is also a solid.
Rubidium is solid at room temperature and reacts violently with water.
Gallium is also solid at room temperature although it would melt at CPU/HS temps. However its cooling wouldn't be any good until it has all melted.
:rolleyes:
Zhaoman
Where is it transfering the heat to? From my understanding the “liquid metal” is just going around in circles keeping the heat within the system?
Yeah, as other replies have said, I think a fan would be needed, which means there are “moving parts”. Doesn't that defeat the purpose?
Brewster0101
What about cost of this very interesting yet new technology.
Yeah, I'd like to get an answer on that.
Their website's quite primitive. Will wait for Hexus' review on this - that should clear things up.
(Edit: Just seen the last post was a while back. Sorry to “thread-urect”.)