[win2kpro]

NOTE: This test WAS NOT conducted in a room where the temperature could be regulated to within + or - 2° and the results below are strictly based on my observations and opinion.

I recently did a test on two thermal compounds I had not used before. They were Innovation Cooling Diamond "7 Carat" Thermal Compound and ARCTIC COOLING MX-2 Thermal Compound.

The Innovation Cooling Diamond "7 Carat" Thermal Compound is nice and thick. I used a heat gun to warm the compound in the tube before applying it to the processor heatspreader. Applied a "dab" about the size of two bb's right in the middle of the heatspreader and let the pressure of locking down the cooler do the spreading. The transfer "footprint" from the processor heatspreader to the heatsink was perfect.

Used the same amount and same procedure for the ARCTIC COOLING MX-2 Thermal Compound. The compound in my opinion is "thin" and the transfer footprint from the processor heatspreader to the heatsink in my opinion was poor.

The processors used in this test were both Intel socket 775 units one being a Pentium 4 631 and the other a Core 2 Duo E6700. In BIOS at idle the Tc temp was approximately 6-7° lower using the Innovation Cooling Diamond "7 Carat" with both processors using the same Intel stock cooler.

Tried an identical test using Arctic Silver 5. Innovation Cooling Diamond "7 Carat" was about 1-2° cooler, however this may have been insignificant since the ambient temperature had changed a little.

After the tests I immediately trashed the ARCTIC COOLING MX-2 Thermal Compound. I very much like the ARCTIC COOLING coolers, however in my opinion their thermal compound leaves something to be desired.

It should be noted that the ARCTIC COOLING MX-2 Thermal Compound has very good reviews at Newegg, but I didn't observe anything to make me want to change from the thermal compounds I presently use.

[hrlow2]

The best test like you were trying would be where ALL things are identical.
Your testing was flawed due to the Thermal differences between the processor types.
Would have possibly been different if CPUs were the same for both compounds.

[win2kpro]

I don't know what you didn't understand, they were identical. I tested the P4 631 with all 3 compounds, and the E6700 with all 3 compounds. Tc temperature variations remained about the same for both processors. Both processors have a TDP of 65W.

As I said in the beginning there was nothing really scientific about this test since I could not regulate the room (ambient) temperature closer than + or - 2° F.

This was just a "quicky" test. Back in 2005 I set up two identical machines with P4 3.0Ghz. Prescotts and used two different thermal compounds and a Honeywell Thermal Pad. In those test we regulated the room to + or - 2° F. and only found the variations to be within the margin or error for the room (ambient) temperature change.

The 2005 tests were done before either of these compounds I used last week were on the market.

Since you seem to question the results, why don't you perform some tests and post your resuts, and maybe you can correct the flaws.

[Stoner]

Were those temps recorded with no load or some level of work being done?
How long did you run each test?

[win2kpro]

Idling in BIOS 5 minutes each.

[loserOlimbs]

Thanks alot Win2kPro

I seem to remember you giving some price points for the different compunds before as well. What gives you the most compund for the least buck out of the Innovation Cooling Diamond "7 Carat and Artic Silver?

If they are close or the same, I might try the Innovation Cooling Diamond "7 Carat next time.

[win2kpro]

1.5 grams of Innovation Cooling Diamond "7 Carat" about $7

3.5 grams Arctic Silver 5 about $8

4 grams Arctic Cooling MX-2 about $8

So 7 Carat is roughly 70% more than AS5. For the results I saw I don't believe I would be inclined to use it to produce a large number of machines. If you are building 1 or 2 and wanted a couple of degrees of bragging rights I guess it would be OK.

I still use some TC-1996 which I like very much. When Shop Intel was in opertion I believe I paid $1 per syringe. It was nice to work with since the exact amount was preloaded in a syringe and you just emptied it on to the processor heatspreader, clamped down the cooler, no measuring or guessing, empty it and throw away the syringe. I believe there was .5g per syringe so it was about the same price as AS5 or MX-2.

The only place I have seen it online since Shop Intel closed was a site where if I recall correctly they wanted about $2.50 per syringe. The temperatures with TC-1996 always ran about the same as AS5.

[hrlow2]

In regards to post,#3, excuse me. From post #1, I assumed thae you used the compounds 1 for each CPU.
I have done an identical test on my machines.Both are IBM8090's with P4 (3.0GHz HT).
Used Arctic Silver 5 on one and Arctic Silver Ceramique on the other.Ran both machines over a week long period while folding on each.Short periods of downtime to allow compounds to cure.
At the end of the week period, the one using Ceramique was running 5C cooler.
Both machines were running side-by-side and got the very same downtimes and everything else were as much the same as I could get it.

[win2kpro]

It's interesting that you mention "cure" times. I've read quite a bit about changes from the compound "curing". This is something I haven't personally seen.

I changed the cooler in the machine I'm on right now a year ago, July 2008. It's a P4 3.0Ghz. Northwood and I used the Intel TC-1996 compound. After I changed the cooler and restarted the machine I checked the processor Tc temperature and it was 37-38°C using the Intel Active Monitor reported temperatures.

I just looked at the temperatures using the same program and the pic is attached. I wonder, since I really don't know; Are cure times applicable to all thermal compounds?

Of course I see temperature variations as room temperatures changes, but it seems (at least on this machine) that after a little more than a year of use the Tc temperature has remained the same at approximately the same room temperature.

[hrlow2]

Cure time didn't seem to make any difference to either of the 2 compounds I tested either. Both at the same temps now.