Maximum PC

ok
I got my tube of IC7 today.

here is the baseline with
4gb of ddr2

running
prime95
folding in the background on a smp client
folding on my gpu radeon 4850
and just generally not shutting anything off while I run prime95

my current temp is rock steady at 53C or 127F





got a few files downloading rite now, vmware program to add a linux client to my folding fun. So, this afternoon when its all done an ready gonna take the pc apart an apply the new thermal paste


I still have to pull an "idle" temp as well but honestly, It always runs in the 115-125 range according to CPUID. Come to think of it I have many other readings from my smp set up night mare let me grab one of them for the "idle" reading lol.



hope this shows a "normal" range before the new paste
certainly gives one an idea of the abuse I give this processor!

ok took apart the whole PC

removed the coolermaster v-8 CPU cooler
removed all the AS5 cleaned it with isopropyl alcohol
nice an shiny

as you can see here on the lower right side you can see the real temp reading of 45C there





thats while running prime95, folding on my gpu client, utorrent etc etc

now apparently real temp 3.0 and HWMonitor are about 5C off from each other. SO if you add 5 to the 45 it would read 50C which would still be an improvement from the baseline





well it's sunday morning now and the paste has had a few days of cure time.
I currently have linux running in vmware and its hitting my cpu steady at 100% load and I am holding a solid 43 C on both cores. I have been running like this now for almost 24 hours an no change, 43 43 on real temp 3.0.

The following having posted their results will get a free tube of ICD24 when the project ends- we are shy of about 15 at this point


tommystarman
MAXPCreader07
guoruiwu1994
Massimo
xethan187x
whathuhitwasntme

My results should be coming up around the middle of May.

Was going to do it this week, but got delayed because of a network hiccup.

At the very least I will have them by the end of this month .

So whats the difference between the ICD7, and the ICD24?

17

My results shall come the day after tomorrow. I wanted to make sure that I had enough TIM left for my new rig before I used it on replacing the stuff in my laptop..

OK Here's the temps again (updated. Too many things could have changed in a month):

Stock Asus Thermal paste:
Core 2 Duo T5800:
Idle: 41 C Load: 59 C

Nvidia 9650m GT:
Idle: 52C Load: 64 C

Now I am going to replace the TIM...

Damn it took a while to get rid of all the old goop. I had to chip it off of the HSF with the edge of a credit card... Heh I guess my folding baked it onto it.

ICD7:
Core 2 Duo T5800:
Idle: 39 C Load: 54 C

Nvidia 9650m GT:
Idle: 48C Load: 56 C

The level of isometric-hexoctahedral crystal lattice structured material

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.