Maximum PC

Here's something from the end side, and is more of an observation than anything. I've been running for a month now just to see something more long run than instant gratification. Just for loads, I was running F@H at 100% and one instance of Prime95.

One brand that I really liked started to go straight to hell. I let it do it's degrading thing, and at exactly one month I pulled the HS's off all the test samples.

Here's what I saw with the OCZ that had early great results and ended up crapping the bed:

It's pretty much mostly been pushed out during the run, as it wasn't like that on it's initial application. It started as mostly a liquid state, and what is left isn't. It's all dried up and there's little to no material left to act as the thermal interface. What started very well, ended up just being bad. I'm going to stop recommending this product in the forums now.

Here's one pull on an ICD7 run


And another ICD7 pull-off:

Looks and acts just as it did on day one. It's still moist where it counts, and the stuff that squished out was like a dry clay.

This was the size I was applying on all brands too:


I had a rough time in one application. I really hated doing it, but had to be tried out. I have no idea how credit card spread users stand doing it, because I couldn't. Felt soooo bad that I was wasting a good tube of ICD7.

What a mess that thing was to clean up! Makes no sense to do this at all.

I have no idea how you got it so messy spreading like that, Chumly (the bottom picture). That's the only way I do it and never get it anywhere but the very top of the heat sink.

Eh, to each their own...

No need to do that though. Using less, and just going right out of the tube works best.

That very amount spreads, as you can see in the above pictures once pulled off.

Here's that amount pressed under glass during a right-up:


Why spread it then? You never have to mess with anything, and it spreads just by putting the HS/F on.

The specs say the the there are the same percentage of diamond in the IC Diamond 24, and the IC Diamond 7. The IC Diamond 24 is 4.8 grams and the IC Diamond is 1.5 grams. If you divide 24 by 4.8, you can see that there are 5 carats per gram of thermal paste. But, if you divide 7 by 1.5 you get 4.67. So that was why i was initially confused.

Ya, sorry...was being a smartass a bit. Big tube vs small tube. Both are 92% Diamond.

Looked at some of the small text on the tube of the IC Diamond 7 Carat, it says that there is actually 7.6 carats in the tube, now it makes much more sense.

Ya, and the big tube says 24 carats, but both state 92% pure diamond by weight.

J finally was able to test out the ICD7. My results are below.

Specs
Intel Core 2 Quad Q6600 2.4GHz
Stock Intel heatsink
2GB DDR2 667 (2x512MB generic, and 1x1GB Corsair XMS2)
Gigabyte GA-P43-DS3L
Generic DVD Burner
eVGA 7800GTX stock clocked
Antec Smartpower 2.0 500w power supply
Windows XP Home Edition with SP3
1x250mm side case fan (intake)
1x80mm intake front case fan
1x120mm rear case fan (exhaust)

Ambient Room Temperature: 24C

Testing Procedure
I loaded two instances of Orthos (prime95 with a GUI) and set the affinity on one instance to two of the processing cores and the other to the remaining cores. This maxed out the processor at 100%. I set it to Small FFT's which stresses the CPU only, for obvious reasons since I don't want to test anything else. I ran the test for 10 minutes, at which time the temps stabilized. I used Core Temp to measure the temps.

Arctic Silver 5 (default paste)
Idle: 52C
Load: 77C

ICD7
Idle: 46C
Load: 73C

how do you install IC7 Like that picture ? i try do it with credit card but won't work.

He shoved it against a glass plate.

Chumly makes a key point here with his tests, if you spend some amount on a compound which has as a key selling point of performance and the performance advantage only lasts 2- 4weeks or 250 - 500 hours? I have a system in my office that's pushing 3 years with IC Diamond and temps have not changed at all since the first week of installation.

It's a two sided equation - Performance & Stability/Reliability

Here are the results to date - If I have missed anybody let me know or if corrections are needed.

yep, i used the stock paste that came with the ocz vendetta.

my base line would be 46/47 C with no load just idling
sorry if I didnt make it clear

Thanks for the clarification I will get it updated

A quick follow up a few weeks later

my temps across the board are rock solid while folding 24/7 I never see them go over 49/50 C in the middle of the day when them ambient temp in the room is 85 F and Higher as we are bumping 94F out side here in south florida

So, I am seeing it improve as it settles in. It was up around 52C Folding and has yet to get there again in the last 2 weeks.


All Temps taken by real temp 3.0

whathuhitwasntme, Thanks for the update. Long term are just as important as performance alone.

We have a few still outstanding?

I am updating all forum results to date for the website 10 forums and you guys are included

I had some time on my hands and felt like I should revisit this project.

I took a quick idle temp, its not a "real" cool down shot, because that would require I actually let the pc set some time. It does however show you the warm idle temp for my pc.





then I re started folding and prime 95 and about everything else I can find to really abuse the processor. Ran it about 30 minutes at 100 % cpu use. Full on max abuse. results as shown here:





I have cupid HARDWARE monitor an REAL TEMP 3.0 running to show a odd temperature opinion from them on the CPU temp. They agree on the GPU at 51C but disagree on the CPU averaging a 5 C difference of temp reading on the two.

I was pretty sure there was no point in hoping that prime 95 was actually going to get much done with 2 instances of folding at home running and sucking 100% of the cpu. It did get a little done but would honestly take a long term run to see it generate a result for prime 95.

quick list of the hardware to compare by

E8500 core 2 duo stock clock never over clocked
GeForce 9800 GT vid card
6gb ddr corsair
raptor 10k hdd not the new ones the old style

oddly enough it was the GPU that heated up more than the rest, seems that inside the case must have been generating more heat so it was taking a hit on temp performance. It typically runs in the 47/48 C range while folding.

The ambient temp was around 76 no special cooling added. All air cooled no water in the case.


this stuff just keeps improving with age for me so far. I have seen it drop another 6 degrees C in baseline temp. Honestly would like to know if the core 2 duo's run in or around that range? I keep reading about i7s being in the low 30C temps and it just amazes me that they running so cool compared to my 45nm wolfdale CPU. Cant wait for the follow up package of stuff to arrive, I have a GPU I need to apply it to!

hmm.. I stopped getting updates from this thread, thought it had died.

I saw some reports recently about the latest i7 not breaking 50C at load...

WHY CAN"T MY i7 RUN THAT COOL!!!!!!


My ICD7 is running great though. Everything is still nice and cool.

Looks like no stragglers left to post. I thought we would have had more results. Many thanks to those that took the time and effort

In any event probably close it out next week and after that for those that posted their results I will make arrangements to send them their free tube of ICD24


Thanks again

Andrew