IC Diamond Test Results - Page 5

Last message on previous page:

Intel has a Flatness spec of .002+/-//(parallel)

This can be some what problematic for people that like to improve their temps and why lapping can sometimes improve temps several degrees.

being flat is most important.




Figure #1 is an idealized contact.

Sink and IHS are perfectly flat. With proper pressure and heat the paste flows and the gap separating sink and IHS closes or resolves down to the average particle size - the size of the particle determines the ultimate bond line thickness (BLT) Smaller the gap the better the performance in conjunction with material thermal conductivity determines performance. My experience is that too much paste will not affect performance -excess is just forced out

Figure # 2 is an illustration of the interference of high spots between the mating surfaces.

If the high spot is the same height or higher as the particle thickness, Particles on top of that high spot will add a whole new layer of compound to the interface, 2X -4x the amount of paste in the BLT increasing the resistance and negatively impacting performance - Flatness is the first priority when seeking best performance.




Flatness again is most important to getting the most out of you system, a max of 600 grit is kind of the limit I have not seen much improvement beyond that as the compound particle size is generally larger than the polishing imperfections.

Although some people really like a good lap job

I like a good lap job...

OK, Here are my results...

Stock heat paste
GPU: idle - 37 C
load - 67 C

IC Diamond 24 Carat
GPU: idle - 34 C
load - 64 C

CPU temp did not change after repasting unfortunately...maybe a bad repasted...
But I will soon repaste the CPU and see if the temps change...

Ok, that makes more sense

Yeah least your temps are gettin down. TBH if you're getting 28-29C with an ambient of 22.2C thats pretty good.

I'm glad we hunkered through your cooling issues and got you happy!

Check your PM instructions provided

I got those tests done... I even did a GPU pressure test, which looked very snug compared to my H50. I don't know what to expect, but with all four tests it looks like the HS wasn't seated right on the H50.

I tightened it as hard as I could without breaking my mobo and did 4 tests on the cpu and 1 on my gpu. (Used all of the test material)

I got something out of it though! I used IC Diamond on my Gigabyte 5870 @900mhz Core

GPU temp before - 45c ~ 62c

GPU temp after - 41c ~ 57c

4-5c drop allover on my 5870.

Apparently I need to do some sanding or something to smooth out my cpu and heatsink surface, but no matter... My temps are good enough.

I could use another H 50 test I have some more contact tests slated to go out next week

Our Method was developed by forum users on the first 4 or 5 giveaways we did. As a group of a couple hundred more or less they tested lines, rice sized, manual spreads, cross etc. The "pea sized" amount turned out to be the best for the high viscosity of our compound.

So we spent some time investigating it to refine the method and explain results and our application instructions page was the result of the process.

This prompted the contact/pressure testing to examine those tests where people do not see a result with the end result we can explain 96% of what goes on.

There are 3 things that determine paste performance

1. Application - Amount and method used.Relatively easy to observe and control.

2. Contact - Flat is best - lap to 600 grit for best performance

3. Pressure - Each compound has their own pressure/performance curve along with different viscosity's. IC diamond is recommend 50 PSI >

If you are trying to define an application amount without the context of 2 and 3 above you are basically shooting blanks in the dark because you have no reference point on the performance curve or on the contact quality.

I believe most people generally understand that testing is a non trivial pursuit so read multiple reviews and multiple user experiences and mentally average to draw there own conclusions.

So in short I highly recommend using our method as proved out in over hundreds of user tests.

Alright, so I'm finally getting around to posting my results.

Gaming laptop P-7801u: I originally wanted to do this one mainly for the graphics card. I then saw just how much disassembled to nothing you have to get to access that heatsink and decided against it. I may one day go back to it. But the CPU and northbridge were accessible from a panel on the bottom of the laptop.

I wasn't aware how important teh thermal pad was for the northbridge and got my fingers all over it and put it down on the counter. I then read the disassembly guide. It said you have to reuse it since normal stuff was unsuitable besides the few very dense stuff including IC Diamond, boy was that lucky.

So:

Ambient: 26°C

Before (Stock paste/pads): Idle: 29°C
Load: 42°C

After: Idle: 29°C
Load: 33°C

That's quite a difference.

I also applied it to my Phenom II X3 720 with a cooler master hyper 212+ in a coolermaster HAF 922 case. Before testing I was running at stock, but decided to overclock a little just to make a bit of a difference. It apparently barely made any difference at all. It's now running at x16 or 3.2 gigahertz per core. Also I used speedfan and the AMD overdrive for reading temps, they still looked lower than ambient, so I guess you'll just have to use the delta between the two. Also, the cooler ran at different speeds so if it had run at the same speeds this new stuff would have done even better.

Ambient:26°C

Before (AS5): Idle:22°C
Load: 31°C Fanspeed at load: 1068RPM

After: Idle: 20°C
Load: 29°C Fanspeed at load: 966RPM

AMD Phenom II x4 @ 3.6ghz
Cooler Master Hyper 212+
Ran two tests of Cinebench 11.5 so all four cores are at 100% performance. Used HWmonitor to capture temps. All the cores where within 1c of each other so I just posted the highest temp.

Pre-IC7
Compound: Antec Silver
http://www.newegg.com/Product/Prod [...] -_-Product
1st run
Idle Temp: 32
Load Temp: 52

2nd run
Idle Temp: 32
Load Temp: 52

Installed IC Diamond as per instructions and let it cure for several hours.
1st run
Idle Temp: 31
Load Temp: 51

2nd Run
Idle Temp: 32
Load Temp: 52

I did notice while playing ME2, my CPU would usually be in the 52-53c range but with IC Diamond its in the 49-51c range. That is just an observation though, not a controlled test.

my fan is set on a manual controller so it can't be changed by the motherboard.

Did you just happen to buy that laptop on Ebay? haha I sold mine, it was a P-7801u(Edit : I think it was 7805) with 9800m gts and 2.8ghz T9600 processor.

Anyways, I just posted a video on youtube showing a small benchmark in gtaiv using ic diamond... cpu was running warmer than normal at idle. Here is a link.

http://www.youtube.com/watch?v=iRWMH7UgdbM

btw, what the $*(# is this above me! LOL

lol. Yeah I was talking about the spam actually. The website title didn't look like it would be selling computer components...eh. DAM HAXXORZ ! lol

hi,
i am john,I have a hitch when removeing my stock heatsink to apply the IC Dioamond, I noticed one of the retaing clips to the heatsink had failed at some point. It was loose and the black part fell into two pices when i unlocked it. I mannaged to get the heatsink back on not to loose to test the same heatsink with the IC Dioamond. Im showing 14degree diffirance in core temp with a 2degree diff in ambient but cant say how much the defective clip effected the results. thank you so much

i also enjoy spamming websites with my failing SEO business so please click on the link that i seem to have lost so that i can get a higher google search rank

This topic has been desticky in top of the forum by Jpishgar

Probably do the final report in beginning of September.

Closing out the giveaway phase tomorrow 12 midnight and will give the last group 30 days to post or have their name entered on the Wall of Shame.

I will post some interesting research this weekend on abrasion.

Hint: What material do they make sandpaper out of and is nearly as hard as diamond?

Lab analysis next week on Darshe's h50 and other loose ends tied up next week as well as updated results

Sorry for the late post, but just finished the last of my exams, Woo-hoo!! I finally got a chance to work with the tube of IC7 I got last week that has been glaring at me during my studies just begging to be used lol

First let me just say that I wanted to thank IC Diamond for providing me with the chance to test out their TIM, and I would definitely recommend this to a friend

Without any further ado, the results are... *drumrolls*

Pre-IC7
Compound: AS5
Abient Temp: 28°C
Idle Temp: 40°C
Load Temp: 69°C

IC7
Ambient Temp: 28°C
Idle Temp: 38°C
Load Temp: 63°C

As you can see, the idle improved by 2°C which although a minor improvement is still a very modest one nonetheless. Wasn't expecting much gain on idle, so any improvement is a welcome one.

Where I was really blown away were the load temps. 6°C cooler using the IC which is roughly an 8.7% improvement over the AS5! It absolutely blew me away that simply changing the TIM could cause such a drastic improvement on performance! I'll never think the same way when it comes to TIM's again lol The only reason I'm sad is I don't have the means to run any more tests, but if this is any indication I'm sure I would not be disappointed!

I have a new build coming up, and will definitely be taking these results to heart when choosing my TIM. Thanks again IC Diamond for providing us with a great product! Keep em coming!

My results:

Arctic Silver 5 + CoolerMaster 212 Hyper+
(AS5 cure time, ~2 weeks)
Ambient ~18C/63F
Idle: 20C
Load: 35C

IC7 + CoolerMaster 212 Hyper+
(IC7 cure time, 2 hours)
Ambient ~18C
Idle: 18C
Load: 31C

Idle temps dropped by 2C
Load Temps dropped by 4C

Seems like good stuff!

Now i see why ICD chose diamonds as a part of their compounds.
"Under normal conditions, diamond has the highest thermal conductivity of all known materials." -link http://en.wikipedia.org/wiki/Carbon
Will return with results.

You can check the heat sink temperature with a laser thermometer. That will give you a basic idea of how high the temperature on the Nb is.

I do not know my NB temp eider, but i know that in really warm days this summer it gets hot, i can not keep my finger on it for more then 4 seconds. It is overvolted and no direct fan on it, but the sys temperature sensor is somewhere else.
If it is just warm then i would not worry about it, if it is how and you can not keep your finger on it for 10s then you can try to change the thermal paste.

My results:

AS5
IDLE
core 0 47C
core 1 43C
fan idle 1600 rpm

Load with prime 95 blend test for 1 our
core 0 66C
core 1 65C
Fan speed 2500rpm

ICD
IDLE
core 0 45C
core 1 42C
Fan speed 1500 rpm

Load with prime 95 blend test for one our
core 0 63C
core 1 61C
Fan speed 2500rpm

Cpu cooler: arctic freezer 7(not rev 2)

Monitored with CPUID HWmonitor, CPU-Z, core speed 3400 MHz from 2.66ghZ on a core 2 duo E6750.
Bus speed 425MHz, rated fsb 1700 MHz, core voltage set in bios to 1,387v, ambient temperature ~28-32C.

No game has ever stressed the CPU this much so i say a blend test is ok, i have never reached over 62C i any game, metro2033, NFS, Crysis, COD..etc.
Thank you for the chance to test this, it has given me lower temperatures. Also i do not know how much pressure my coolers exerts. The test was done after 24+ ores.

PS: the ICD test i think was made on higher ambient temperatures then AS5

Motherboard died 2 days before IC Diamond arrived

All I'm left with now is my laptop. Don't think I will be able to provide any results in a timely fasion, as it will most likely be a month or two before I get a working desktop pc again

CPU: Pentium 4 (630 SL7Z9)
MoBo: P965T-A
RAM: 3 GB Corsair DDR2 533
HD: Seagate 7200.7 160 GB (ST3160827AS)
GFX: ATI Radeon X1600 PRO (stock heat sink)
CPU Cooler: Stock Intel heat sink

Used Hardware Monitor to get temps, Prime95 and 3D Mark06 trial version for load testing.
Room temp between 20-21C
All recorded temps where taken after 2 hours operation for their prospective tests.

Manually set Intel heat sink to 100%
P4 Idle / Load Arctic Silver 5 (6+ months cure)
46 / 61 C

P4 Idle / Load IC Diamond (~6 hours cure)
43 / 58 C

A cool 3C drop in idle and load.

Forgot to Find a program that would manually set the X1600 fan speed, I was to excited to take it apart.
X1600 Idle / Load Stock Paste (3+ years cure)
54 / 73 C

X1600 Idle / Load IC Diamond (~6 hours cure)
52 / 71 C
~10% / ~20% Less Fan Speed

So a definite 2C cooler + a quieter fan for the X1600.