News - Danamics LMX breathes new life into liquid-metal CPU coolers

[HEXUS]

Successor to last year's pricey LM10 promises to make liquid-metal cooling more affordable.

Read more.

[Main]

Danamics mk2 eh, it certainly looks more refined than the first one.

You aiming to get one into the lab? It would certainly be interesting to see if it does any better than the Corsair H50, and that wasn't even around when they first tried to bring their liquid metal cooler to the market.

[JCBeastie]

If I remember right the reason these didn't sell isn't solely down to the ludicrous pricetag but more to do with the performance, while Danamics claimed it was more efficient than 'water cooling' all real world reviews showed performed reasonably close to Thermalrights Ultra120 and other air-coolers that cost less than half as much, demonstrating that heatpipe technology is a far cheaper and more elegant solution.

I don't believe the liquid metal cooling technology is ever going to be either as efficient or cost effective in a stand alone format as conventional coolers, maybe a chassis incorporating it to cool the major components but not a HSF.

[Funkstar]

I don't believe the liquid metal cooling technology is ever going to be either as efficient or cost effective in a stand alone format as conventional coolers, maybe a chassis incorporating it to cool the major components but not a HSF.

I wouldn't say it is that cut and dry. This is only the second iteration of the product, I would expect performance to increase with each release. Regular heat pipe coolers have been around for years now, with many design improvements and changes along the way, increasing performance at the same time. I would expect the same from liquid metal solutions, especially if other manufacturers introduce their own variations.

If only Dynamics have products like this then obviously their competitors don't see the technology having any traction.... or they have it tied up very tightly with patents.

[JCBeastie]

I wouldn't say it is that cut and dry. This is only the second iteration of the product, I would expect performance to increase with each release. Regular heat pipe coolers have been around for years now, with many design improvements and changes along the way, increasing performance at the same time. I would expect the same from liquid metal solutions, especially if other manufacturers introduce their own variations.

If only Dynamics have products like this then obviously their competitors don't see the technology having any traction.... or they have it tied up very tightly with patents.

I'm not just making arbitrary proclimations, the issue is that the form-factor is ultimately limited by surface area, it doesn't matter if the liquid metal is more efficient at moving heat from the CPU at the end of the day it's still aluminium/copper fins that are dumping that heat into the air, and that'll happen at the same rate as an equivalent heatpipe tower which is cheaper to manufacture. Exactly the same reason why tower type water-cooler units failed.

[Funkstar]

I'm not just making arbitrary proclimations, the issue is that the form-factor is ultimately limited by surface area, it doesn't matter if the liquid metal is more efficient at moving heat from the CPU at the end of the day it's still aluminium/copper fins that are dumping that heat into the air, and that'll happen at the same rate as an equivalent heatpipe tower which is cheaper to manufacture. Exactly the same reason why tower type water-cooler units failed.

That's actually a really good point, something I hadn't considered either

If that is the case, then however clever this technology is, it is ultimately pointless.

[shaithis]

I'm not just making arbitrary proclimations, the issue is that the form-factor is ultimately limited by surface area, it doesn't matter if the liquid metal is more efficient at moving heat from the CPU at the end of the day it's still aluminium/copper fins that are dumping that heat into the air, and that'll happen at the same rate as an equivalent heatpipe tower which is cheaper to manufacture. Exactly the same reason why tower type water-cooler units failed.

Isn't that assuming that the fins are already disipatting heat as fast as they possibly can?

I have no idea if modern heatsinks are pushing the fins to their limits or not.....if they aren't, then getting the heat to them faster would be advantageous.

[Main]

I'm not just making arbitrary proclimations,

Sorry, but you are

the issue is that the form-factor is ultimately limited by surface area, it doesn't matter if the liquid metal is more efficient at moving heat from the CPU at the end of the day it's still aluminium/copper fins that are dumping that heat into the air,

Ummmm, no, it does matter. It matters a great deal. Surface area is only one part of heat transfer.

If the means of shifting heat from the chip to the fin-stack is good enough then you can do a better job with the same, or even with less surface area. The Sapphire 3870 Atomic/Toxic, with it's single-slot vapour-chamber cooler which beats larger coolers for performance is an example of that. If you can do better with less area, you have the possibility of making a cheaper cooler because you use less metal, even if the core technology is more expensive. In fact vapour chambers are being used on the 5770's stock heatsink. That's high-end novelty to stock cooler in a few years, not bad for a new technology.

Another problem is you're assuming the technology isn't going to mature. Heatpipes were nothing but a gimick when they first appeared on PC coolers. I know this because I had one of the first ones, and it was crap. Liquid metal cooling may not have had a successful application in the PC cooling market yet but that doesn't mean there isn't potential for it to be. It's quite possible there are already solutions to similar problems out there in obscure industrial/military applications that aren't far from what we have here. Danamics could easily be a commercial venture for a company that makes such things.

Heatpipes themselves also have operating limitations, in that the amount of heat they can shift is dictated by the temperature heat source and heat sink ends are at. Both this cooler and water coolers use a pump to move the heat transfer medium at a rate independent of the operating temperature and so performance can be increased simply by improving the pump. In fact, performance increases were seen in reviews of the original Danamics cooler when more power was supplied to the pump. I wouldn't be suprised if this new cooler beat everything out there, but if it did, yes it would still be a case of whether the performance would be worth the cost.

...and that'll happen at the same rate as an equivalent heatpipe tower which is cheaper to manufacture.

Yeah, but the rate at which a cooler shifts heat is not what is used to measure a cooler's performance because that rate is fixed by the chip being cooled. If a chip produces 150W at full load, then, assuming all that heat is dumped by the cooler, any cooler attached to it, regardless of design, will dump that amount of heat (assuming the operating temperature is below the chip's thermal cut off). The performance is judged by the temperature the chip must run at for the cooler to do its job.

Exactly the same reason why tower type water-cooler units failed.

If they failed because a cheaper cooler can shift the same heat load, then everybody would be using stock coolers because they're cheapest.......

But yeah, I get what you meant to say, that combined water coolers failed becasue they didn't give enough perfromance to justify the price, and that could quite easily happen here. Again.

[baius]

Would also like to see a review.HEXUS @ some point. Hexus' reviews are typically excellent, so I'll be looking forward to it...

[schiffsratte]

Well , already the LM10 kept my 125w 9850 Phenom at 27ºC under heavy load. Never seen more then 34ºc with it at all - so the new one sounds promising anyway