• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

FRONTPAGE Intel i7 3770K - Ivy Bridge - CPU Review

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
So basically ideally we wouldn't see any TIM on the die/heatspreader area then if it was making perfect/near-perfect contact?

Yes this is why i asked, i would bow to Bobnova's expertise.

My thinking was the TIM is purely a conduit to the heat spreader / cooler, i would have thought the more TIM there is between them the more it actually acts like a barrier.

I use AS5 and very thinly, just a very fine transparent layer and i get great temps like that on a really crappy cooler.

My thinking was if the HS was hard pressed against the DIE cap with only a fine layer of TIM in the way it would conduct heat more efficiently, thus temps might improve (if you can make it a tighter fit)
 
It looks like Intel is not doing the TIM thing quite right IMHO.

Optimally TIM is only supposed to fill in the microscopic surface imperfections that cause gaps between two surfaces making heat transfer possible in the gaps via the TIM but still letting the two surfaces touch where they can. Too much and you have the TIM acting as another barrier to heat transfer that would otherwise occur where the surfaces "line up" and touch.

a1f1art.gif <-- White here is TIM. And this application looks to thick.

From TIM article here: http://www.electronics-cooling.com/1996/09/thermal-interface-materials-2/

A technical paper on TIM testing: http://arxiv.org/ftp/arxiv/papers/0709/0709.1849.pdf
 
Last edited:
Functionally, it doesn't matter if the TIM is ideal or not in application.

A solder joint instead of TIM is many times more effective at conducting heat.

By using TIM, they have effectively turned the "heatspreader" into a heat barrier. The conduction of heat would be far more effective if it was die->TIM->heatsink. Die->TIM->heatspreader->TIM->heatsink is bad for heat transfer... same as we found on older chips that used TIM rather than solder under the IHS.

If the paste is applied right or not is only a matter of varying degrees of bad. It cant be good even if applied well.

Article forthcoming summarizing the problem.
 
Functionally, it doesn't matter if the TIM is ideal or not in application.

A solder joint instead of TIM is many times more effective at conducting heat.

By using TIM, they have effectively turned the "heatspreader" into a heat barrier. The conduction of heat would be far more effective if it was die->TIM->heatsink. Die->TIM->heatspreader->TIM->heatsink is bad for heat transfer... same as we found on older chips that used TIM rather than solder under the IHS.

If the paste is applied right or not is only a matter of varying degrees of bad. It cant be good even if applied well.
I agree. You said it better than me.:attn:
 
We'd see a very thin fairly translucent layer, rather than the thick opaque one shown.
Unless the photographer cleaned up the processor first, it almost looks like a solidified mass that only adhered to the IHS as it was peeled off.
 
Nice article. About the TIM between the heatspreader and core... I thought it was some sort of heat conducting epoxy?

Guess I'll hold on to my Nehalem i7 920 a bit longer until Intel releases a more OC friendly CPU.
 
Ace's test comforts me. The released batch must be hit or miss on temp. Retail version hopefully will be check for quality control. They got to be with now including warranties for overclocking.
 
He's on water and stock v/undervolted. His temps I do not believe to reflect more common water setups and especially air. Here is to hoping that is right though!
 
One would think Intel could fix the TIM issue very quickly if they see it as an issue. Once the initial batch is released they can change to their proprietary substance that they used in SB and we'd all be happy. Whether or not that will actually happen...
 
It's not a matter of changing substances, but of changing types of interface. Solder is much better than TIM for conductivity and is more than a mere changing of substance as far as implementation.
 
Back