Hardware Canucks

Hardware Canucks (http://www.hardwarecanucks.com/forum/)
-   Air Cooling (http://www.hardwarecanucks.com/forum/air-cooling/)
-   -   Lets talk about the future of air cooling (http://www.hardwarecanucks.com/forum/air-cooling/37897-lets-talk-about-future-air-cooling.html)

BootlegUsher November 11, 2010 10:44 PM

Lets talk about the future of air cooling
 
So it seems like to get better temperatures using air cooling, manufacturers are making bigger and bigger heatsinks. Each new design might improve by 1-2 degrees, and not much is happening. After reading the GTX 580 review and how they are using this vapour chamber design, it seems like it is much more effective then the previous heatpipe based cooling solution that was on the 480. Wouldn't this make sense for CPU heatsinks? Will we ever see the Thermalright Ultra 120 Vapour Chamber Extreme?

Thoughts?

enaberif November 11, 2010 10:50 PM

Personally I'd like to see a big movement in air cooling, but with self contained units like the H70 and others offered by Coolit its pretty tough to compete in that market.

Most air coolers have a TDP that CPUs saturate fairly quick with crazy overclocks.

_dangtx_ November 11, 2010 11:15 PM

well, if they take the hr-02 and add a vapour chamber? sweet jesus.

i ghettorigged a huge arse half messed up jobbe once, the rad was massive, the whole lenght of the case, and it was all passive. i used small heatpip[es though, and i doubt much was circulating. thats what i see. massive area, no fans, or something inbetween...like the side of the case being the cooler itself, like in some notebooks where the whole plate on the bottom is a heatsink.

Mook06 November 11, 2010 11:59 PM

yeah thats the first thing i thought when i saw the 580 heatsink, i wanna see someone mod and slap that on a cpu

the tdp is similar, no? 150w for the 480, similar for 580?

edit: wow i was way off, the 480 is around 275w, 580 is 244w

i really wonder how that heatsink would perform on a cpu

DCCV44.2223 November 12, 2010 12:36 PM

Not sure if vapour chamber can work for CPU because of the limited space (horizontally) around the socket, it has to go up first and then spread out.

Zalman did heatpipe fanless case a few years ago, TNN500 and TNN300.

::: Zalman, leading the world of Quiet Computing Solutions :::

I think the TNN 500 had a list price of something like US$1200?

MpG December 19, 2010 10:59 PM

When it comes down to it, vapor chamber and heat pipes really aren't that different in the first place. Both rely on a phase change cycle within an artificially-created sealed environment to move heat away from a source faster than just conduction could manage. It's just a different shape of chamber and form factor involved. Video cards don't have much height to work with, but lots of surface area, so it makes sense to create a wide surface area to attach the fines to. Whereas CPU cooler have clearance limitations, so it makes sense to go up to an more open area.

The biggest advancement at this point would probably be some new base material (i.e. carbon nanotubes), or a CPU IHS with a larger surface area that had a wider spread of heat, which would allow for more heatsink contact area. Of course, the latter would be solely up to the chip manufacturer.

CMetaphor December 19, 2010 11:37 PM

I want the liquid metal heatsink to be legal for sale in canada.... sigh, lol.

ilya December 20, 2010 01:04 AM

Quote:

Originally Posted by MpG (Post 464550)
When it comes down to it, vapor chamber and heat pipes really aren't that different in the first place. Both rely on a phase change cycle within an artificially-created sealed environment to move heat away from a source faster than just conduction could manage. It's just a different shape of chamber and form factor involved. Video cards don't have much height to work with, but lots of surface area, so it makes sense to create a wide surface area to attach the fines to. Whereas CPU cooler have clearance limitations, so it makes sense to go up to an more open area.

The biggest advancement at this point would probably be some new base material (i.e. carbon nanotubes), or a CPU IHS with a larger surface area that had a wider spread of heat, which would allow for more heatsink contact area. Of course, the latter would be solely up to the chip manufacturer.

This.

Quote:

Originally Posted by CMetaphor (Post 464559)
I want the liquid metal heatsink to be legal for sale in canada.... sigh, lol.

I think it's technically legal for sale in Canada, just illegal and next to impossible to ship/import it here, not to mention benchmarks don't show much of an improvement over the high end air coolers. Plus you'd have to worry about the electromagnet failing and causing a catastrophic meltdown inside your system.

Back on topic, I would really like to see Zalman get back into the game and get some inspiration from their competitors. The way I see it, their designs have the most potential compared to the big guns like Noctua and Thermalright. One thing I've always liked about their "flower" style heatsinks/towers was the fact that you could mount them in any direction without worrying about the heatpipes having to fight gravity. (huge issue with my Prolimatech Meghalems)
To improve their heatpipes, I thought of creating something like a closed heat "loop." Then twist it into one of the classic Zalman "omega" or "infinity" shapes. Doing so in a way that the heatpipes always use gravity to their advantage, no matter what the position. Capillary forces on expensive sintered heatpipes just aren't good enough IMO.

Another company I would like to see stay in the game is Prolimatech. The way their heatsink is designed is just pure genius. Most companies stagger their heatpipes and bend their fins to get more heat exchange, Prolimatech appears to simply use air pressure. (with their original Megahalems design) The dead zone in the centre (which is sealed off) seems to make very good use of Bernoulli's Principle. More air pressure means more heat exchange with less turbulence and therefore in theory should result in a more effective heatsink with less noise. Kind of like the concept of the Dyson bladeless fan. I'd really like to see more companies put these concepts to use. Too many companies seem to be ignoring basic physics and just trying to create a marketable product ASAP to meet demand, or simply go with trial and error. There are so many smart people out there that can work out theoretical designs, put them to good use!

The third innovative element I'd like to see more of is the spiral airflow stuff. (no clue what to call it) It's the stuff in the MSI cyclone, newer Intel stock HSFs, and the Silverstone Air Penetrator fans. Sweeping fan blades create turbulence, but the spiral behind it sort of counters it out and increases static pressure, which in theory should also increase heat exchange at a lower noise ratio.

So my dream HSF would probably involve a "spiral flower tower" with a low pressure zone directly behind the fan hub and "omega" or "infinity" shaped heatpipes to prevent loss of efficiency when fighting gravity. Add cool materials that we normally don't use in HSFs and I'll be happy.

I've also realized my paint skills are horrible, so I'm not posting a pic.

CMetaphor December 20, 2010 07:57 AM

Quote:

Originally Posted by ilya (Post 464571)
I think it's technically legal for sale in Canada, just illegal and next to impossible to ship/import it here, not to mention benchmarks don't show much of an improvement over the high end air coolers. Plus you'd have to worry about the electromagnet failing and causing a catastrophic meltdown inside your system.

Back on topic[...]

Its not off topic because its technically still an air cooler of sorts... And I would worry much less about an electromagnetic pump with no moving parts failing than a water pump with moving parts.
Also, the reason it isn't legal is because the liquid metal fluid they use is slightly dangerous and not yet approved by the CSA.

Moreso on topic? Heatpipes and Alu/Cu have pretty much reached their limit. They can't get much better or heavier anymore - theyre already huge and heavy in some cases. A new technology will be required in the future to deal with the increases heat loads.... because I'm pretty certain that the CPU makers of the world won't always stick to a 125/140w TDP limit, but who knows?

SKYMTL December 20, 2010 08:56 AM

People seem to forget that a vapor chamber and heatpipes are mutually exclusive. The second you add heatpipes to a VC design, any benefits derived from having the fins, condensation chamber and processor in direct contact with one another are completely eliminated. Heatpipes just introduce the same bottlenecks into the equation as before.

Personally, I don't think we will ever see a vapor chamber-equipped heatsink for enthusiast or "gamer"-branded setups. There is actually a simple reason for this. As I stated above, for a vapor chamber to work as advertised there needs to be a direct and short path between the item emitting heat and the cooling array (in this case the fins). That isn't all that hard but in order to disperse all of the heat built up within the vapor chamber a large surface area of fins is needed. This doesn't mean height as we have seen with modern heatsinks but rather WIDTH is needed to maximinze the number of fins for efficient heat distribution.

This total surface area which is VERY close to the CPU causes an issue: motherboard VRM and chipset heatsink designs not to mention memory locations don't allow for much space between the CPU socket and an area of potential conflict. This is why heatpipes are used. They extend the fin array up and over any "higher" areas on the board but at the same time cause a bottleneck in the cooling process.

Vapor chambers can be used on GPUs and in laptops because both have low-slung components. There is simply no way to make a vapor chamber heatsink that is designed in such a way that it somehow avoids the frankly idiotic cooling features on every enthusiast motherboard.


All times are GMT -7. The time now is 03:14 AM.