Uber cooling foil?

linky

eh? Their website doesn't mention much about it.....and t here's no direct quotes, nothing like "the lap techs at RNT say blah blah,"
but the article sounds convincing.....

seems like a pretty good idea too me...

it sound like a good idea. it would bring temps down fast.

Very Intriguing.. I'm hoping that this isn't just something you think will work in theory, but works just as well in practice. (And more importantly that this article wasn't paid for by RNT!!)

/me Waits Patiently scratching chin

I think ultimately the best application for this would be at fab-time.

AMD would have to move to an integrated heat spreader.

Once that decision was made, the core could be flash-welded to a fairly large heatspreader, with excellent thermal transfer.

This is a very interesting development, I hope the Big 2 move on it when it becomes a marketed product.

Edit:

It also occurred to me as I hit submit that GPUs could easily benefit from this as well, by flashwelding the core to a heatspreader of industry-standard size. Then just about any properly sized hsf could be put on, depending on the company's (and end-users) whim.

Ok. I emailed him yesterday about it... I aint eveng ot to read the link LoL! I knew it'd be here in like 5 min of it being posted...

Down to biz...

The simple fact is that this isn't going to solve anything... The heatsink is going to be getting more heat. It's not going to be enough of a change from what we have now to drop temps much.

The way he goes on about it he makes it seem as if we'll find some "holy paste" that will drop everyone's temps 15C. Yeah right.

Heatsinks now are once again dealing with heatloads that are (or will be) comming close enough to what they can deal with while remaining "tolerable" in the noise and cost.
^ Keep in mind that's pretty much in the worst case.

My point is that I really don't see how we're going to be dropping temps a ton from where they are now. Obviously this is like saying who needs more than 640kb of ram. I just am more inclinded to think even a huge leap isn't going to do much in a relative since.

I mean it seems every version of AS get's 1-2C better. So that's 6C better than white grease I still pull out for some things.

Yea! <sarcasim implied>

BTW heat spreaders don't help or hurt any if all right now in any relevant level.

That and the problem is still the HEAT->Heatsink transfer Be that by a core, or the core ->heatspreader->heatsink.

This will make no large difference. The improvement has to be made between heatsink and CPU directly... If you don't eliminate or improve the TIM joint then you aren't really improving anything. And you can't reasonably bond a cpu to a heatsink in mass production.

Sure if you use this with a heatspreader it might improve that bond, but heatspreaders just introduce another layer for heat to migrate across and are not a cooling improvement - they're CPU insurance.

I hope nothing like this is ever used by a major cpu manufacturer!

Think about it: the heatsink being bonded to the core?!?!?! Unless we were to take huge risks, that would only leave us with changing fans FROM A STOCK AMD OR INTEL HEATSINK!
Ugh.

Fortunately, I don't think cpu's are a valid market for this technology- mosfets, IC chips, amny other things, but not cpus.
Hopefully.

rogerdugans said:

I hope nothing like this is ever used by a major cpu manufacturer!

Think about it: the heatsink being bonded to the core?!?!?! Unless we were to take huge risks, that would only leave us with changing fans FROM A STOCK AMD OR INTEL HEATSINK!
Ugh.

Fortunately, I don't think cpu's are a valid market for this technology- mosfets, IC chips, amny other things, but not cpus.
Hopefully.

Click to expand...

yeah i agree but i think most ppl were trying to say that this is a good way to increase heat transfer between the core and the heatspreader, this would kind of create a bigger surface area for the chip because the heat transfers so well from the core to the heatspreader and we would'nt have to worry about trying to cool the tiny cores as hard.

max

Lt. Max said:

yeah i agree but i think most ppl were trying to say that this is a good way to increase heat transfer between the core and the heatspreader, this would kind of create a bigger surface area for the chip because the heat transfers so well from the core to the heatspreader and we would'nt have to worry about trying to cool the tiny cores as hard.

max

Click to expand...

I see this possibility...

I wonder how well heat would conduct out across a larger heatspreader?

It would have to conduct across the heatspreader well enough that the advantage of larger surface area of the heatspreader was greater than the disadvantage of increasing the number of layers between heatsink and core.

My point is this:
What makes a heatspreader superior over the base of a heatsink at conducting heat? And then, is the heatspreader enough of an improvement to also alleviate the negative effect of increaseing the distance heat has to travel and the resistivity of conduction across the different layers?

There is still virtually the same surface area of contact between the core and the heatspreader as there would be with the core and the heatsink... It's hard for me to see how a heatspreader could really ever be a considerable improvement when it comes to heat transfer.

well if you use that foil to put the heatspreader on the cpu, it has 10x the thermal conductivity of any paste, so the heat should have no problem conducting all throughout the heatspreader. the question is, how good a thermal conductivity is in the processor itself? this way we can compare how close thermal conductivity of the foil is to the thermal conductivity of the processor itself. right now the heat transfers from the core to the heatspreader with an inefficient thermal paste which probably is not very good at evening out the heat over a bigger area, but it works fairly well.
its just the smaller cores that make it harder for the small amount of thermal paste to transfer the heat to the big heatsinks that this could be a very good idea in theory at least.

max

IMOG said:

My point is this:
What makes a heatspreader superior over the base of a heatsink at conducting heat? And then, is the heatspreader enough of an improvement to also alleviate the negative effect of increaseing the distance heat has to travel and the resistivity of conduction across the different layers?

There is still virtually the same surface area of contact between the core and the heatspreader as there would be with the core and the heatsink... It's hard for me to see how a heatspreader could really ever be a considerable improvement when it comes to heat transfer.

Click to expand...

i have to agree with IMOG, rogerdugans, and Toysrme. Even if the heat gets transferred better how much of a temp difference will you see? And if processors start shipping with "permanently" bonded heatsinks, thats not going to be a good thing.

Lt. Max said:

well if you use that foil to put the heatspreader on the cpu, it has 10x the thermal conductivity of any paste, so the heat should have no problem conducting all throughout the heatspreader. the question is, how good a thermal conductivity is in the processor itself? this way we can compare how close thermal conductivity of the foil is to the thermal conductivity of the processor itself. right now the heat transfers from the core to the heatspreader with an inefficient thermal paste which probably is not very good at evening out the heat over a bigger area, but it works fairly well.
its just the smaller cores that make it harder for the small amount of thermal paste to transfer the heat to the big heatsinks that this could be a very good idea in theory at least.

max

Click to expand...

Thanks for the dialogue Lt. I see your point.

If this foil was good enough to alleviate the disadvantages of heatspreaders as they are, then I think this would be a great thing too. You'd get good heat transfer, plus chip cracking/crushing insurance. Sounds like a good deal to me.

After the clarification by Lt. Max, I do see the possible benefits, as IMOG said....

I use hose clamps on tight tube connections in water cooled systems as insurance; if heatspreader efficiency could be improved to the point that material claims, I can't say I'd mind having core-crush insurance