News Sapphire brings Vapor-X tech to CPU cooling

[Gareth Halfacree]

Vaporisation cooling offers 200W of dissipation.
http://www.bit-tech.net/news/hardwar...-vapor-x-cpu/1

[Tyrmot]

Review incoming?

[GuilleAcoustic]

Interesting ... waiting for reviews !

[Bloodburgers]

Sadly i am not in the market for a new cooler until i re-build my current system however for those who are they are on offers at ocuk. (i do not work for ocuk i hasten to add)

http://www.overclockers.co.uk/showpr...odid=HS-000-SP

[flibblesan]

I need to see a review on this as these type of CPU coolers have been tried before and failed.

[Bakes]

Can we stop with the whole bit-tech 'can dissipate x heat'? Of course it can dissipate 200W heat; anything can dissipate 200W heat. Seriously, if a pinhead were at a sufficient starting temperature it would be able to dissipate 200W heat at room temperature, because the temperature gradient would be high. Even the Artic Cooling Freezer 7 Pro was marketed at being able to dissipate 200W heat, back in like 2007.

[Roskoken]

Originally Posted by Bakes

Can we stop with the whole bit-tech 'can dissipate x heat'? Of course it can dissipate 200W heat; anything can dissipate 200W heat. Seriously, if a pinhead were at a sufficient starting temperature it would be able to dissipate 200W heat at room temperature, because the temperature gradient would be high. Even the Artic Cooling Freezer 7 Pro was marketed at being able to dissipate 200W heat, back in like 2007.

To think google have been spending a fortune on elaborate server cooling for all this time when all they needed were some pins!

You should email them RIGHT NOW with your theory.

[foxrena]

Originally Posted by Roskoken

To think google have been spending a fortune on elaborate server cooling for all this time when all they needed were some pins!

You should email them RIGHT NOW with your theory.

It's funny everytime I see comments like yours. You think you know but you don't.

This is simple physics... claiming 200W heat dissipation is meaningless if no temperature difference is given. It's just a way of advertising for the company. I would hesitate buy such a product if I see these misleading ad numbers.

What truely matters for a cooler is the C/W ratings, as should be used by all cooling enginners. Anything can dissipate 200W, @Bakes is right. The stock cooler can dissipate 200W of power when it's 30C higher than ambient while a high performance cooler dissipates the same amount with maybe only 10C higher than ambient. When you compare 0.15C/W with 0.05C/W, you know which is better.

Sapphire engineerers know this. But their marketing department do not care.. They will just throw out a numbe .... 200W, sounds plenty, Ha!

[feedayeen]

Originally Posted by Roskoken

To think google have been spending a fortune on elaborate server cooling for all this time when all they needed were some pins!

You should email them RIGHT NOW with your theory.

What Bakes said is physically accurate. Newton's Law of Cooling says that heat transfer is directly proportional to both surface area and temperature difference. If you make either one of these values large enough, it compensates for the other.

[xxxsonic1971]

Looks like Darth Vader built it....

[Roskoken]

Originally Posted by feedayeen

What Bakes said is physically accurate. Newton's Law of Cooling says that heat transfer is directly proportional to both surface area and temperature difference. If you make either one of these values large enough, it compensates for the other.

Only if its linear

[feedayeen]

Originally Posted by Roskoken

Only if its linear

I figured you would have dropped this. Newton's Law of Cooling applies to the lumped capacitance model of thermal bodies where temperature remains at a constant temperature inside both bodies. This improves the rate of heatflow since the gradient is so narrow. For small or metalic objects, it's a good approximation, for air, not so much since the thermal conductivity sucks. But even when you factor this in, you wind up with a temperature curve that looks like 1-e^-x with some constants to scale the amplitude so that x=0 produces the temperature of the heatsink and another to determine the half life.... but even this is linear since the derivative at x=0 is proportional to the temperature.

I've already determined that it's linear so I could stop this right now, but it's actually better than that. If you get really, really hot, well above what semiconductors can survive at, maybe 500-1000C depending on geometry, the warm air will have significantly less density than cool air, convection has now become a major force.... In practice, it reduces the size of this half life and thus makes the derivative at the boundary x=0 larger, thus more heat flow.

You then have blackbody radiation, which is proportional to 1/(e^(1/x) - 1), this curve starts out less than linear and it's one of the least efficient modes of heat transfer, but it accelerates slowly.

[Bakes]

Originally Posted by Roskoken

Only if its linear

Shh you. Even if we completely ignore Newton and just stick the pinhead computer in space, by the time we reach 10000C it's emitting about 500W in black body radiation alone. If you can make your components survive temperatures of 10000C, then by all means attach a pinhead for cooling purposes, because it'll get the job done.

[Kris]

I love the discussion - who knew you could get from a CPU cooler to black body radiation with only so few comments.

In all seriousness though: what if sapphire actually, against all odds, took the number 200W as it's "normal" mode of operation and performance as compared to the competition? (like keeping a 3960X at 50C while it is OC-d to 4,5GHz or whatever) One can hope.

[Gareth Halfacree]

Originally Posted by Bakes

Can we stop with the whole bit-tech 'can dissipate x heat'? Of course it can dissipate 200W heat; anything can dissipate 200W heat.

If Sapphire had released any other figures, you'd have those too. As it is, "can dissipate 200W" is as close as the company has got to talking about performance. If it makes you feel any better, assume a worst-case scenario of "can dissipate 200W of heat without allowing the CPU temperature to raise to a point where it would trigger the chip's thermal cut-out."

In other words: yes, a pin head could dissipate 200W of heat - but not without raising the CPU temperature to a level at which it stops operating. The Vapor-X CPU - and any other cooling system that says "can dissipate XW" - by contrast can.

[LennyRhys]

My GPU uses a similarly hyped "vapour chamber" style cooling and although it works well, it's still a bit of a gimmick. Heatsinks do what they are meant to do: they take heat away from the component and dump it.

If anything, this heatsink has arguably missed the bus by going the 120mm route; the very best air coolers are all 140mm and it's unlikely that a 120mm heatsink will be able to compete, whether it uses vapourisation or not.

EDIT: the "dense aluminium fin stack" is also a bit of a step backwards... all it means is that the cooler will need powerful fans for optimal performance (or, conversely, that it will fail dismally when furnished with quiet fans).

EDIT 2: The American spelling of Vapour is really bugging me.

[jamsand]

Looks like it'll be loud as hell, 40 at 2200rpm!!! I'll stick with my silverarrow or d-14

[Gareth Halfacree]

Originally Posted by LennyRhys

EDIT 2: The American spelling of Vapour is really bugging me.

Same. It's a proper noun, sadly, so I have to override my fingers every time I try to 'correct' it - i.e. every single time I type it...

[Bakes]

Originally Posted by Gareth Halfacree

If Sapphire had released any other figures, you'd have those too. As it is, "can dissipate 200W" is as close as the company has got to talking about performance. If it makes you feel any better, assume a worst-case scenario of "can dissipate 200W of heat without allowing the CPU temperature to raise to a point where it would trigger the chip's thermal cut-out."

In other words: yes, a pin head could dissipate 200W of heat - but not without raising the CPU temperature to a level at which it stops operating. The Vapor-X CPU - and any other cooling system that says "can dissipate XW" - by contrast can.

Yes, it's all you've got, but it's a totally useless piece of information, because it tells us nothing about the quality of the cooler. For example, a standard 3770k will throw out about 77W when not overclocked; basic electronics tells us that if at 5GHz with a fairly dramatic 1.45V vCore it will be sticking out about 148W - which is enough for an Artic Cooling Freezer 7 Pro v2, which is rated at 150W (sorry, didn't mean 200w in the previous post).

Would you want to run those settings with a Freezer 7 Pro? Probably not.

All the numbers are saying is that what they've come up with is at least a viable CPU cooler. < Not interesting info.

There's not really any point attempting to quantify it/buying into the marketing info; the information means nothing, because any modern processor would be running at obscene speeds before it would actually release that much heat (I may be overlooking some of the Sandy Bridge-E parts here). < my point

[Harlequin]

Originally Posted by jamsand

Looks like it'll be loud as hell, 40 at 2200rpm!!! I'll stick with my silverarrow or d-14

silverarrow is noisy as heck compared to my passive hyper Z600