Amazing ICD7 results

Hello everyone. I received some Innovation Cooling Diamond thermal compound and have done some basic short term tests on it so far. Initially I was pretty skeptical about how a thermal paste could possible change temperatures that much compared to others. I have even tried Arctic Silver 5 and the generic white thermal paste and saw absolutely no differences in temperature changes.

Here are my initial results and will update them possibly a month or two down the road to see if this thermal paste can still perform the same as when it started.

*All results were taken at 23C ambient with the same humidity.



I used some kiddie site to make that graph to better show the differences.

Here are the same results in text format:

BEFORE

CPU IDLE - 38C 38C
GPU IDLE - 62C

CPU LOAD - 46C 46C
GPU LOAD - 77C

AFTER ICD7 APPLICATION

CPU IDLE - 32C 32C
GPU IDLE - 51C

CPU LOAD - 40C 40C
GPU LOAD - 66C

These results are the immediate results with practically no cure time on my M1330 with Copper insert mod. Idle temperatures were with the laptop with fan off and load temperatures are with fan at about 3000rpm.

Overall I am extremely pleased with these results immediately after application. I observed a 11C and 6C drop in temperatures from GPU and CPU, respectively. The temperature drops were the same on load AND idle which is very impressive to me.

I will be applying this to all my systems and will post more results as I finish the testing. Expect Latitude D430, D420 and XT results to be coming shortly.

EDIT

To those who are skeptical of my results I don't blame you. The original paste has actually been reapplied about 3 or 4 times yielding about the same results before. And I know not to get any oils or dust particles on the paste or core itself. I use rubbing alcohol 91% and q-tips to wipe away old paste and dust particles in between applications.

 

Nice work bro !

 

unless you re-test with other heat goop formulations and then again with ICD7, you have no 'control' over the beginning conditions of your test.

The initial temps could easily have been caused by bad/incorrect application of the original heat goop/pad. Your careful installation of ICD7 may have simply fixed the bad application.

 

those are impressive results!

 

If you are that impressed by ICD7, you will be even more impressed by the several thermal pastes that are better than it.

 

Which are better? I've used Arctic Silver 5, but it squishes out the sides after few months. I was planning on buying Arctic Cooling MX-3 next.

 

if you are squishing heat goop out the sides of a heat sink, then you've likely applied too much

 

I know there's that one Japanese thermal compound with the very interesting name. Supposed to be the best but rare and expensive.

 

Shin-etsu is the best, but not easy to work with.

I prefer MX-3 and some others here will tell you OCZ Freeze. Either are very good

 

No, there wasn't any where it should have been; the die was almost bare. I used the right amount but it must not have cured very well.

 

either that or the die/heat sink had finger oils on it.

finger oils even in minute quantities are the kiss of death to most heat goops.

 

How much are we supposed to put on? I'm not getting nearly close to those results. We're supposed to put a pea sized diameter in the center and put the heatsink on am I correct? The pressure from the heatsink installation should "squish" it out enough to properly cover the cpu?

Lmk fellas!

 

pea size? no...much smaller than that. Think more like a grain of rice, maybe a bit bigger depending on the size of the surface. This is what I've been recommended and it's worked fine for me.

 

It's supposed ti be like that. The goal is to have as little TIM as necessary to fill the minute, including microscopic, air pockets between the die and heatsink, not form some boundary layer between the two.

For the past year or so I've been spreading the TIM over the die with a razor blade so that it completely but very thinly covers the entire die. I've noticed at least a degree difference with every application. Also when I remove the heatsink there is always a much thinner layer of TIM between the die and heatsink than if I just put a small blob in the center. Also the only thermal paste I have been using for the past 6 years has been from the same giant tube of arctic silver ceramique. I have a tube of the good shin etsu stuff but it will probably be another year before I get to it!

 

I know it's not a boundary layer. I fully understand what it's purpose is. When I say nearly bare, I mean most of the die didn't have any thermal paste on it.

 

While you are at it, I recommend using brasso to polish the heatsink contact surface to remove copper oxide layer will improve thermal transfer.
Of Course after polishing it you need to remove all traces of brasso using a non-conductive solvent like isopropyl alcohol.

 

If you've got bare die, it's very likely you've got reduced thermal conductivity. I second advocating the razor blade / credit card method - enough to fill gaps, get total coverage, and avoid "squishing." I probably use, at most, a drop the size of half a BB to cover any size heatsink and die (this really just means the die and roughly the contact patch on the heatsink though)