Another Observation Confirming Diamond TIM Performance

[BonzaiDuck]

Some may remember a long thread last summer about motherboard ducting, diamond thermal paste, and the ThermalRight Ultra-120-Extreme. I had determined through my own tests -- producing bar-charts of temperature reading frequencies -- that IC Diamond or a mixture with similar particle loading can reduce temperatures on an over-clocked E6600 or Q6600 by 2 to 3C degrees.

Today I reseated an SI-120 cooler on a Prescott 3.4E processor (TDP=105W). The cooler had previously been installed with Arctic Silver 5 TIM. I used IC Diamond this time.

The idle temperature at today's room ambient with Arctic Silver has always been 4C higher than it is with the diamond TIM, and the load temperature seems to be correspondingly lower by the same 4C degrees.

 

[PCTC2]

Thanks BD. I just ordered IC DIAMOND for my new watercooling setup for my Q6600 and then also for my Thermalright Ultra120 on my E6400. I had AS5 on it but I had to reseat the U120 on my Q6600 without replacement AS5. Now I can't wait to see the temps of my Q6600 with my ApogeeGTX. Woo.

 

[aigomorla]

#$!@$#!%$!# how the hell do you apply that thing correctly like that to see that big of a difference.

This thing is a serious PITA to apply correctly.

 

[georgeszi]

My experience with AS5 is the the temp drops a degree or so first day or so but after a month or so they end up a degree or two higher than the initial install(after 2 hours).

I think that there is some pump out on that stuff so an aged install of AS5 against a fresh application you might see a little higher spread. The 40-50 people that have tested ICD7 @ overclockers.com seem to get larger differences on the old install vs. fresh applications. Not that ICD7 doesn't seem pretty good as it seems to beat AS5 by a pretty good margin on fresh applications as well as older ones and is supposed to be formulated to protect against pump out.

ShinEtsu and MX2 seem also to fit the bill as far as stability I have had good luck with those also vs AS5. Although ICD7 seems a little better from user results.

Looks to me on the top greases it seems easy to get performance but the real question is how long they hold or maintain that performance

 

[GoodRevrnd]

Bonzai, did you do an even coat, a bebe in the middle, or a cross?