Wednesday, May 4th 2005

NanoCoolers puts liquid metal in your PC

Most of you may not have heard of NanoCoolers yet. This company is a relatively young Startup(May 2002) who has been working on improving cooling technologies behind closed doors for quite some time.

Now the first details on their cool new products emerge:
The liquid metal has significant advantages over other single phase liquid solutions. The thermal and physical properties of the material give it the ability to cool extremely high heat fluxes. With its very low vapor pressure, the boiling point of the material is in excess of 2000°C. This provides the capability to cool extremely high power densities without the liquid-metal changing phase, removing power density as the limiting factor in cooling performance. The liquid metal is non-flammable, non-toxic and environmentally friendly. As a metal, the liquid is both highly thermally conductive and highly electrically conductive. The thermal conductivity makes it ideal for heat removal and dissipation. The electrical conductivity enables the use of electromagnetic pumps to propel the liquid.
We got our hands on a 16 page confidential presentation:

Add your own comment

86 Comments on NanoCoolers puts liquid metal in your PC

#1
gR3iF
hmpf okay here in long but in this case my english might not be enough.
u know these things from the old days wich you get when you are ill and you were too hot?
inside there was a metal this thing is the only that is fluid while beeing on room temp

so this is the only known metal that can be used for this cooler
and this metal is toxic really toxic
Posted on Reply
#2
Unregistered
wazzledoozle
I see 2 fans :rolleyes:
It says "requires no moving parts" it didn't say it "has no moving parts."
Posted on Edit | Reply
#3
Unregistered
gR3iF
hmpf okay here in long but in this case my english might not be enough.
u know these things from the old days wich you get when you are ill and you were too hot?
inside there was a metal this thing is the only that is fluid while beeing on room temp

so this is the only known metal that can be used for this cooler
and this metal is toxic really toxic
http://www.lenntech.com/Periodic-chart-elements/melting-point.htm

Francium has not yet been made in a stable manner yet, but both cesium and gallium are liquid metals at just above room temperature, and certainly at "chip" temperature. In addition, the reference I gave to the patent earlier, a patent applied for by nanoCoolers, the users of this 'liquid metal', clearly states, amongst other things, a predominantly gallium-indium alloy "One such preferred composition comprises 66% gallium, 20% indium, 11% tin, 1% copper, 1% zinc and 1% bismuth".

In fact "The various properties of Ga--In alloy make it desirable liquid metal for use in heat spreaders. The Ga--In alloy spans a wide range of temperature with high thermal and electrical conductivities. It has melting points ranging from -15.degree. C. to 30.degree. C. and does not form vapor at least upto 2000.degree. C. It is not toxic and is relatively cheap."

This clearly states a metal alloy which is non-toxic, relatively cheap, high boiling point and low melting point (albeit not as low a Mercury) metal. Stop saying you don't know, and just read the post, properly this time.
Posted on Edit | Reply
#4
Unregistered
Unregistered
Believe me -- No matter what versions, it does need fans to take heat to the air! Liquid metal is just used to take heat from the chip!
Yes, but the use of the liquid metal does allow for a much more varied design of heatsink. I agree that fans are practically a necessity (my information of no fans is based on their press release) but it is possible to use much better and larger fans, which produce less noise for a given air flowrate. Coupled with a more efficient heat exchanger design, this should help to reduce the temperature of the chip quite substantially, even if large, low noise fans are still necessary.

In addition, I would like to point you to this information:

"Some early testing showed that prototype cooled Radeon X850XT PE card to just 12 Celsius.

It can work completely fanless and can cool the high end fastest and latest greatest cards. Sapphire will have this product ready in July and will launch many other SKUs based on it."

http://www.theinquirer.net/?article=23331

Now, given that I have a Radeon 9800XT and, despite the nice copper heat exchanger and fan on the unit and the two fans in my case, I find it a challenge to get the temperature below 75deg.C (without a cool breeze from outside), I think 12deg.C is pretty impressive, and, even without the fans, I'd imagine that the temperature would still be pretty good. As a final note, you could even simply run the fans you do have on the card, but at a lower speed to reduce the noise, taking advantage of the superior heat transfer properties (or at least their current claims).

As a second final note, if you check their (nanoCoolers) presentation, you will note that they only specify no moving parts in the pump (obviously in contrast to water cooled systems). This shows that they did not specify the lack of fans on the final system. Quite frankly, I think in the end, few people who are willing to buy this system would be prepared to turn off the fans to save noise in preference to showing off to their friends that their liquid metal cooling system kicked the seven shades of ..... out of their water cooling system, even given the noise.
Posted on Edit | Reply
#5
Unregistered
tinkerer, phd student

Looks exciting. My guess is that the "liquid metal" is in fact some sort of metal suspension (or possibly solution) in another liquid (ever notice the magnet on some oil filters? It's there to catch metal particles from engine wear). I'd put money on any metal being liquid at room temperature as being toxic or too damn hot to safely use in a home or office. As for Nuclear reactors, I grew up next to one... and the "cooling" being done is actually part of the energy transfer loop - all industrial US reactors use good ol' steam engines powered by steam from a secondary or tertiary loop (to prevent radioactive contamination). The 'coolant' is supersaturated steam in the primary loop which vaporizes the water in the second loop to drive the steam engine. Using "liquid metal" in a reactor coolant scheme is not quite what it sounds. Nuclear reactors all work by converting the simple heat of a moderated nuclear reation in to mechanical energy and then in to electrical energy (save for those in satellites). Water will always be in the primary loop. In the secondary loop they are using a liquid metal scheme which converts the heat energy in to motion of the liquid metal which then creates an electromagnetic field and thus electricity... the heat is then re-caputured from the liquid metal, probably by water again, and either drives another steam engine or simply dissipates to atmosphere (or the ocean) as non-radioactive steam (or just warm water). It all comes down to efficiency - how much heat energy you can convert to mechanical and then to electrical. When you're producing megawatt electricity, every portion of a percent counts.

As a side note Heavy water is used as a moderator in the nuclear reaction itself (so is lowering and raising control rods). Funny thing is, plants are designed so that they can only be active when there is a certain amount of heavy water present to "moderate" the neutron speed. Also, pebble bed reactors simply represent a cheaper design, safer . more fail-safe, fuel form (encapsulated "pebble" fuel), and a more efficient energy transfer scheme.

The slides we all looked at were clearly intended for a non-technical audience. My bet is that it was for a venture-capital audience or business management-level people in the company.

I also noticed a slide about thermoelectrics... which is, by the way, a horrible way to cool something because it produces a great deal of excess heat. I don't know the details, but I know they can get very hot on the back side. My guess is that the company is planning on sandwhiching a thermoelectric sheet between the cpu and the heat sink. This would increase the heat transfer rate from the chip and possibly cool it below room temperature (which you cannot do with a conventional H20 based liquid system, but would have great ramifications for home-pc performance). Since the liquid metal (whatever it is) has a high heat capacity and high thermoconductivity, it may heat up well beyond what would be traditionally expected of gpu or cpu chip and thus explain much higher temperatures. Given Newton's law of cooling - it may actually be a good idea to have the liquid metal run hot so long as the temperature of the thermocouple on the cpu/gpu side is cool. This will allow increased thermal conductivity to the radiator blades (especially if there is limited surface area) and an increased temperature differential between the blades and the ambient air temperature. However, I think many people are missing the point... I've seen some discussion of CPU temperatures which doesn't make any sense unless you consider at what rate it is being cooled (watts). For example, you can have a 1000 watt lightbulb be cool to the touch if it's being cooled (heat is being taken away somewhere else) fast enough OR, a 25 watt bulb will burn you to a crisp if it's enclosed and the heat has no where to dissipate. So, in the end, I'm sure whoever is putting this "liquid metal" cooling scheme together is finding an optimal solution given the standard thermodynamic equations.

One thing I really like, however, about their solution : no condensation! If you were to use a refrigerant-based system (like in a typical house)... which is easy to make small - it would be hard to keep the pipes going to your chip from dripping water on everything else ... bad stuff.

As for the magnets - magnetic shielding is trivial and moving "liquid metal" at the low power consumption they are talking about will probably produce a far weaker field than the circuits on your video card.

Well, I apologize for the lack of effort, misspelling, and probable misuse of terminology. I usually don't post to discussion forums.

I would, however, expect this technology to be rather ubiquitous as the power consumption (and thus heat dissipation) of gpu/cpu chips increases.

For the mean time, consider that the more you have to cool your cpu, the more you heat the room it is in... the more you heat the room it is in, the more you have to cool your room, the more you have to cool your room, the more you have to heat up something else.... the air outside. Can't beat the laws of thermodynamics!!!

Anyway, I liked the comment about US Navy... being an ex-sailor.
Posted on Edit | Reply
#6
Unregistered
Unregistered
Not to belittle Canada any more than she already is...but the safest nuclear reactor title? Try the United States Navy. All those reactors on carriers and submarines, being operated by people as young as 21? No accidents? No nuclear-related deaths over the course of 50+ years? Sorry Canada.
LOLOLOL no incidents.. If there even be an accident u think they'll tell it to public?..
Posted on Edit | Reply
#7
Unregistered
"deutronium" utilized as a neutron moderator or radiation shield - ridiculous!

when i read that inane nonsense i almost fell off my chair laughing.

DEUTERIUM is the heavier, non-radioactive isotope of hydrogen, not "deutronium" - that hogwash term came from "lost in space", a goofy sci-fi show I watched when I was a kid in the 1960's.

good lord, how sophomoric.
#8
Unregistered
John at BU
One thing I really like, however, about their solution : no condensation! If you were to use a refrigerant-based system (like in a typical house)... which is easy to make small - it would be hard to keep the pipes going to your chip from dripping water on everything else ... bad stuff.
There's an easy way to keep a refrigerated system cool, but above zero degrees centigrade: limit the duty cycle with a thermostat. Nobody said you had to keep it on all the time.
#9
Unregistered
John at BU
I'd put money on any metal being liquid at room temperature as being toxic or too damn hot to safely use in a home or office.
How can something be too damn hot in room temperature, I asks ya? Is a cup full of 25C liquid metal hot-hot-hot or not?
#10
Unregistered
[QUOTE=Vorlaplex]

In addition, I would like to point you to this information:

"Some early testing showed that prototype cooled Radeon X850XT PE card to just 12 Celsius.


yeah probably in north pole cos liquid cooling cannot cool the cpu below roomtemp.[quote]Not to belittle Canada any more than she already is...but the safest nuclear reactor title? Try the United States Navy. All those reactors on carriers and submarines, being operated by people as young as 21? No accidents? No nuclear-related deaths over the course of 50+ years? Sorry Canada.
--------------------------------------------------------------------------------

LOLOLOL no incidents.. If there even be an accident u think they'll tell it to public?..

Yeah and what about those two "nuclear reactors" that usa dopped to japan? Few hundrend thousand dead! Oh wait that wasn't an accident.
#11
Unregistered
Melting point for this liquid metal is - 20 C . I might add that it is not cheap either you can pick it up for approx. $24 US for 2cc.
Posted on Edit | Reply
#12
Unregistered
This is pretty exciting tech actually.

It is using all the basic principles of water cooling, but it's more efficient and safer. It is more durable and requires little or no maintenance.

For the examples of sub-ambient cooling, they would have to be using the thin-film thermo-electric coolers. In the thin-film version, these devices use much less power and create much less heat added to the total heat burden. I use water cooling, have used TECs all the way up to the most powerfull commercially available for a PC, 437watts. The thin-film thermo-electric TECs give this technology a greater range of application then the Liquid Metal cooling alone, but they are not required for all cooling applications.

As for needing fans, that fully depends on the heat exchanger design. There is no reason that passive cooling design won't work for the final stage heat exchanger. But there are requirements and they would mean a lot of surface area would have to be used for passive cooling to work.

Every chip has a range of temperatures in which they can function properly, the ideal solution is a cooling device which can maintain the chip's temp within this range in normal environmental conditions. For the average PC, this range isn't so bad, but some PCs must be operated in harsh environments, some people are overclockers and induce additional stress changing the requirements. Reguardless, any cooling solution which is simple, requires little maintenance, low cost, and is effective across a wide range of operating envirnments is a good thing.
Posted on Edit | Reply
#13
Unregistered
Not exciting at all.

Liquid metal cooling is just a hoax. Although thermal conductivity is very high, the heat capacity of the liquid metal is so low (365 W/Kg.K vs 4218 W/kg.K for water), that means it need much higher volume flow rate for the cooling. So, it will have almost same performance as a water cold plate, which is much cheaper. Moreover, a MHD pump has very low capability for pumping, although it is claimed as no moving part inside.

As to a TEC, anyone knows its efficiency? Itis around 10% (regular mechanic refrigiration system is around 3~4!). That means if you want to cool down 100W, you need put 1,000 W electriicty there. Anyone is Crazy like that?

Besides, liquid metal never be safer that water. It is corrisve and nobody know its stability yet.



Unregistered
This is pretty exciting tech actually.

It is using all the basic principles of water cooling, but it's more efficient and safer. It is more durable and requires little or no maintenance.

For the examples of sub-ambient cooling, they would have to be using the thin-film thermo-electric coolers. In the thin-film version, these devices use much less power and create much less heat added to the total heat burden. I use water cooling, have used TECs all the way up to the most powerfull commercially available for a PC, 437watts. The thin-film thermo-electric TECs give this technology a greater range of application then the Liquid Metal cooling alone, but they are not required for all cooling applications.

As for needing fans, that fully depends on the heat exchanger design. There is no reason that passive cooling design won't work for the final stage heat exchanger. But there are requirements and they would mean a lot of surface area would have to be used for passive cooling to work.

Every chip has a range of temperatures in which they can function properly, the ideal solution is a cooling device which can maintain the chip's temp within this range in normal environmental conditions. For the average PC, this range isn't so bad, but some PCs must be operated in harsh environments, some people are overclockers and induce additional stress changing the requirements. Reguardless, any cooling solution which is simple, requires little maintenance, low cost, and is effective across a wide range of operating envirnments is a good thing.
Posted on Edit | Reply
#14
Unregistered
Unregistered
oughta pull plug first ;o)!
People like this guy shouldn't be allowed to breed. I say leave the darn thing plugged in !
Posted on Edit | Reply
#15
Unregistered
Unregistered
Not to belittle Canada any more than she already is...but the safest nuclear reactor title? Try the United States Navy. All those reactors on carriers and submarines, being operated by people as young as 21? No accidents? No nuclear-related deaths over the course of 50+ years? Sorry Canada.
I hate to be the one to inform you that the US is not without nuclear screw-ups. Three mile island ring any bells?
#16
Unregistered
-------------
As to a TEC, anyone knows its efficiency? Itis around 10% (regular mechanic refrigiration system is around 3~4!). That means if you want to cool down 100W, you need put 1,000 W electriicty there. Anyone is Crazy like that
-------------
your a moron.
#17
Unregistered
One Question... Why did they pull the pages??

Anyone notice that the company pulled all their pages about liquid metal cooling. Is it possible they jumped the gun somewhere being that they're a "startup" and all. Quite odd.

I remember visiting the site a while back and they listed "Liquid Metal Cooling" on their technology page.
#18
Unregistered
Wow, you guys are pretty hard on each other here. Anyhow, I have it from a reliable source, ie someone who works for Sapphire, that the Blizzard project has been scrapped. It is my understanding that Sapphire wanted to continue with the project but the company who was to produce the heatsinks wanted to go a different direction.
#19
Unregistered
No moving parts means it uses a solide state pump. Basically, if that gallium alloy flows along Z axis, there's aan electric field created along X, and a magnetic filed along Y.

Who said it needs 0.9 T. It's pumpin' metal, not air or salted water, that's not a MHD glider.
#20
Unregistered
Unregistered
Not to belittle Canada any more than she already is...but the safest nuclear reactor title? Try the United States Navy. All those reactors on carriers and submarines, being operated by people as young as 21? No accidents? No nuclear-related deaths over the course of 50+ years? Sorry Canada.
Just a little fact, the United States stopped manufacturing nuclear reactors due to a cost efficency budget, it was found more resourceful to purchace reactors and matenece for those reactors from their neighbour to the north. all reactors used in the states are the Candue reactor, manufacutred in Ontario Canada.
#21
Unregistered
15th Warlock
Next time you're going to call ppl names, get your facts straight, 'cause right now, the above statement makes a 4th grader look ten times as smart as you are...
Have you ever heard of Mercury? it's a liquid metal at room temperature, it's what's inside those old glass things your mom used to put in your mouth (and in other places... :p ) to meassure your temperature. Does the word thermometer ring a bell?
Now, Gallium on the other hand melts at aprox. 29.76°C, well under the idle temp of that thing called CPU inside your PC, but of course you've never heard of gallium before, have you? :confused:
Now, boiling, put in layman terms, means turning any material into gasseus state, (meaning it is not liquid or solid anymore) so, as you kindly pointed out before:
"Unlike water, the metal boils at 2000°C, which means it can absorb more heat without changing phase and becoming a troublesome gas."
Let me translate that last paragraph to you, this means that this specific liquid metal (or alloy) won't turn into gasseus state unless it hits 2000°C, which is about 20 times over the boiling point of water (aprox 100°C depending on the altitude), did you get that, sport?
You sound a right d*ckhead.
#22
Unregistered
So many of you people in here are so far 'up yourselves' it's actually quite amusing. Who honestly cares about a few spelling mistakes apart from an anal retentive hermit ?

Grow up and stop trying to score points off each other...it's like a kindergarten in here.
#23
15th Warlock
Aitch
You sound a right d*ckhead.
Hello, and welcome to 4 months ago, "d*ckhead" :)
Posted on Reply
#24
Unregistered
------------------------------------------------------------------------------------------------
After you know the story, you know you are really a moron.

I met the guy, U. Ghoshal, from nanocoolers in a meeting of June 2005. He talked about his TEC and showed the efficiency is 10%, currently. He agreed that currently nobody can use it until the ZT>=3.0. However, nobody knows when it will come becuase if you reach ZT=2.0, you could publish a paper in Nature.

So, come on. Don't fight each other. Just call the nanocoolers or email to ask some professional questions.

Another one is on MHD pump, the number of 0.9T is from a journal paper. The MHD pump for liquid metal circulation is not new, people study it for a long time. However, if the outer magnetic field is not strong enough, the liquid cannot flow.

In general, my conclusion is the liquid metal cooled video card WILL NOT happen. Ask some experts from thermal engineering to get some knowledge before shouting around here!
-----------------------------------------------------------------------------------------------
Unregistered
-------------
As to a TEC, anyone knows its efficiency? Itis around 10% (regular mechanic refrigiration system is around 3~4!). That means if you want to cool down 100W, you need put 1,000 W electriicty there. Anyone is Crazy like that
-------------
your a moron.
Add your own comment