IC Diamond Test Results

Page 4 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
@4ryan6 -- you are correct that AS5 has about a 200 hour cure time which is partially why they are testing this product against already installed applications of old thermal paste ( ie. testing your current system with cured paste and then removing and applying the ICD paste) instead of having users test new applications of both - so that the cure time for the AS5 has already been reached and then testing with newly applied (or after the 2 hour cure time for it) ICD paste. which makes the results based on properly cured products.

 
Correct, ICD24 is expensive, however a smaller tube (7 carat), is $7. It depends on the user whether or not they think that the few extra degrees lower is worth the price premium. For high end users, most likely, but for average users, AS5 or Ceramique would do.

From the amount in the tube, I think it is worth the price. Though, I've had a tube of Ceramique for almost two years now, reapplying and upgrading constantly, lasted for at least 20 uses.
 
There's a lot of people who pay $150+ for a quality ventilated case, $50-70 for the CPU cooler, and another $50-70 for a GPU cooler. If they have liquid cooling, that's another $100+ for the radiator, block, resevoir, fittings, pump... Adding another $25 for a thermal paste that cools an additional 3-5°C doesn't seem that unreasonable. And since the tube will last 3-4 applications, that doesn't seem bad at all. Hell, that's possibly the most effective use of money per temp drop in everything you buy.
Click to expand...

I agree with everything you said except the cussing, lets try to keep that kind of thing out of any discussions about this OK.

You're not addressing someone, that's never spent a chunk of money trying to get the best results they can, but in this day and time money is tight for everyone.

So $25.00 extra dollars can be a deal breaker added on top of an already large purchase, I'm just asking is it your opinion its worth it, as you guys were all given the testing TIM, you didn't actually pay for it, or did you pay for it?

After looking over your test results with your "Generic Silver Thermal Paste", you didn't actually pay for that either did you, it probably tagged along with your heat sink?
 


The only place I could actually find it that I'm comfortable doing business with was FrozenCPU, hopefully places like Newegg and such as that will eventually be carrying it, Newegg is great about letting you add a 3.5grm of AS5 to a purchase for like 5 bucks or so, and sometimes that's convenient.

The Shin Etsu X23-7762 also out performs AS5 and is also expensive for the amount you get vs price comparison.

http://www.crazypc.com/products/50118.html

Were any of the tests done here directly comparing that product to the ICD24?
 



USA best price if you do a search $17.99 for a 24 carat tube generally 19.99 is the high.

Other Comparison factors to consider

Paste should be compared on volume - 3.5 gm silver weight for silver vs 1.5 grams diamond weight has an equal volume about 1/2 cc

IC Diamond is thick but for a reason, creamy or more liquid type pastes provide great performance out of the gate but due to their low viscosity tend to be more prone pump out or bake out and often performance is limited to just a few hundred hours depending on use. The new ThermalRight compound is recommended to change on an annual basis.

I have a guy(123Bob) on xtremesytems folding 2 years 24/7 overclocked without having to change his settings, his temp did not change during this period

I have an office system 3 years with no change in temp under normal use.

one users experience


http://www.maximumpc.com/forums/viewtopic.php?t=90505&postdays=0&postorder=asc&highlight=diamond&start=48



IC diamond was specifically designed for for high reliability and for long term use. OEM's typically spec bulk loadings above 90% simply because it is harder to pump out a solid than a liquid, pump out being one of the primary causes of TIM failure.

This is a quality concept/issue that I have been trying to introduce to the retail market.

Performance short term is relatively easy to design for as seen in tests like using toothpaste and other household products.

Designing for reliability as you see in some of the OEM compounds is also relatively easy but they fall short in performance (as far as the hobbyist goes)

What is most difficult is to combine high performance and high reliability into one product.

So there is a long term savings cost effect that's not readily apparent by not having to change your compound as often or the time and effort to do so.
 


Something is wrong with the Maximum PC link, but it looks configured right?

I appreciate the information, and was already sold anyway, I had been considering the Shin Etsu X23-7762 but price wise its more expensive per volume than the ICD24, performance wise AS5 has been steadily falling behind, so I was looking for a replacement TIM for my high end machines.

For the performance gain the price is worth it IMO.

Thank You! Ryan
 
Holy Crap! Big tube!

I did the GPUs in my 2nd/folding@home machine.
Ph2 720, Zaalman 9500a, 2X 8800GT, shitty OCZ PSU in shitty NCIX promo case.

Temps are for 24/7 folding, 33C case temps:

Stock: 89C

MX-2: 87C

ICD: 81C after 7 days!!!

Mother of god, you guys are awesome.
 
@ryan (if I daresay), Shin-Etsu X32-7783D (not 63D) outperforms everything except for indigo xtreme.

I'm definitely not baking my CPU at 100C to melt the indigo, and most likely not spending on a TIM that's $10 per gram... :|
 
^^Exactly. The thing is I have a little problem with running my CPU at 100 degrees C at more than a few minutes.

IMO, if you're into pushing hardware, ceramique is the way to go because it can withstand subzero 😛

BTW, @IC Diamond, are you guys planning on releasing a paste that can handle subzero temps? :)
 
The TIM should be adequately spread via the downforce of the CPU cooler (a sufficient amount of downforce will provide best results; minimum 50psi of downforce should be the goal for optimum results. please see the Innovation Cooling website for more info). Also, for coolers/surfaces that have many peaks and valleys, it may be beneficial to wet the surface(s) with a small amount of IC7 to initially fill the voids in the mounting surfaces...then proceed with recommended application as usual. Please click on the link below for application methodology and shoot me a PM if you have any questions. FYI, read the review on this site for Gilgamesh's results and ideas for application



Mercedes Benz 300CE Parts
 


The issue with paste at sub zero and the reason some are favored is a thin vs thick viscosity.

Liquid type compounds resolve down to their BLT or bond line thickness(thickness of the average particle size in the joint) very fast, higher or more viscous compounds take a little more time.

We have guy on EVGA we are sponsoring with sub zero and has success using a variant of the Indigo method of heating the paste up to 90C under clamping pressure and thinning the joint or paste thickness that way.

I have worked out a "cold cure"

I used a simulated copper IHS lapped to about a 600 grit.

Clamped it to a glass slide with some paste and used a micrometer at timed intervals to measure the thickness

@ around 65F ambient within the first min or so .005 - after 5 min .0025. - after 30 min .001. - after 2 hours .0005.

This was at @ 65F ambient with good pressure of about 55 psi.

2 hours gets you about 90% overnight or 4-8 hours will net you another 1/2 - 1 degree C.

depending on your pressure your fast cure method using heat will definitely accelerate the process but would be a function heat, pressure, time. right now I cannot tell you how fast 90 C will cure but my guess is you would need more than a couple of minutes for a full cure. I just do not know whats happening here and have not figured a way to investigate it

Those that start the system say within the first hour or two from an ambient mount to subzero will likely see the compound harden at a thicker level than you would want for optimum results, it really needs to get to it's BLT or average particle size.
 



Shin Etsu is definitely one of the more competitive compounds though rarely used as I only have a handful of comparisons over 400 tests I would like to get my user sample size up to 30 or 40 to get a better picture.



Shin%20Etsu%20august%2012%20users.PNG



This is the user hierarchy established from approx 400 tests, some of the newer compounds were not available at the time the tests were run. I would expect that within a year or so you will see more of a grouping at the performance end. New paste we are introducing is a degree or two less than ICD but we expect to introduce a 4 gm tube @ a $6.99 MSRP Same high reliability, targeted to the value users, the volume user's like the whitebox guy's


Condensed%20Multiforum%20Tests%20-%20391User%20Results.PNG

 
Some late results since I'm probably the last one =P

**STOCK everything (volts, speed, multiplier, NB, SB, etc...**

Processor: AMD Phenom IIx4 955
Cooler: Zalman CNPS 9700
Motherboard: ASUS M4A79T Deluxe
Case (airflow): Raidmax Aztec

I went ahead and included my case and my motherboard since the Zalman 9700 cooler mounts via snap-on brackets and honestly, I'm kinda skeptical about the strength of those. Second of all I mentioned my case since in my setup, there is a 80mm Fan right above (or next to, i should say) the cooler blowing air into it, and then a 80CFM fan that i installed for exhaust and a large fan above the cooler. Some pictures of the setup:

IMG_9502.JPG


IMG_9506.JPG


Also, normally I would have my processor overclocked to 3.7Ghz with some fancy voltage number =P but to keep everything standard i just set everything in my bios to "Auto" which is basically just the stock setup that it comes with.




Below are the test results and all tests had an ambient temp of about 27.2 Degrees Celsius. Also i based all my results off of CoreTemp. Realtemp doesn't support my processor and Speccy and Hardware monitor is there just to verify the temperature just in case CoreTemp is somewhat inaccurate in sensing the CPU temps.



Zalman ST-G1 results after 2 Hours of In Place large FFT's in Prime 95 for maximum heat, then ~20 min cool-down for measuring idle temp: (the generic thermal paste that came with the 9700 cooler)

Load temp: 64 (Celsius)
MAX load temp: 65

load_temp_zalman_paste.png


Idle temp: 48
LOW idle temp: 47

idle_temp_zalman_paste.png




IC Diamond results after 4 Hours (FOUR HOURS, two hours longer than Zalman Test) of In Place large FFT's in Prime 95 for maximum heat, then ~20 min cool-down for measuring idle temp:

Load temp: 57 (Celsius)
MAX load temp: 59

load_temp_ic_diamond.png


Idle Temp: 44
LOW idle temp: 43

idle_temp_ic_diamond.png

**note that in this picture, the max temperature recorded by HW-Monitor and CoreTemp has a significant variance. This is because the min/max values of HW-monitor was cleared out right after testing.
***once again, all values are based upon CoreTemp





CONCLUSION:

It is amazing to see how much temperature difference there can be just by swapping out the thermal grease that goes on the bottom of the Heatsink.

Difference LOAD temp: 7
Difference MAX LOAD temp: 6

Difference IDLE temp: 4
Difference LOW IDLE temp: 4

So in the end, I would definitely use IC Diamond for future builds and even maybe swap out the thermal paste in my graphics card for IC Diamond. Once I'm finished with my current tube, I would definitely go buy IC Diamond although maybe the smaller version since this tube is big and will probably last a good 5 builds. I personally don't have any experience with Arctic Silver 5 or The Ceramique versions or any other versions of that for that matter but from what other people had concluded, I'm gonna go play it safe and go with IC Diamond. 😉

One more thing: Running everything stock on this cooler makes it possible to stay under 60 Degrees Celsius under full load. 60 Degrees Celsius just happens to be the max temperature published on AMD's website (apparently they changed it recently to 62 degrees, but it was rated @ 60 degrees @ the time of my purchase)

http://products.amd.com/en-us/DesktopCPUDetail.aspx?id=532&f1=&f2=&f3=&f4=&f5=&f6=&f7=&f8=&f9=&f10=&f11=&f12=

My Processor is the C2 stepping by the way. And I think with a stock cooler, the temperatures will definitely go above 62 degrees or 60 degrees.
But I personally don't care about this number and sometimes I pushed it past 67 degrees. (W/O IC Diamond of course haha) but for the people who really pay attention to the published 62 degrees, this thermal paste does contribute to keeping the temperature below that. (my theory is that AMD changed it from 60 to 62 because the boxed stock cooler couldn't keep the temperatures below that number so they changed it, and that makes me think that "62 degrees Celsius" is a made up number) This number is also not the "Official" TJ Max for the processor since the official TJ Max is Unpublished.
 
I have a Dell PC with really shotty stuff. The heatsink mounts with screws, not spring loaded, and it made applying difficult. Keeping even pressure while screwing each one slightly in turns left less then confident mounting. I have no idea what Tim they used in the factory installation but after being only 1½ years old, it was obviously cheap stuff. Already dryed out and flakey. I seen AWESOME results replacing the old and inferior material! Stock Heatsink is garbage lol.

Somthing to note as well, once I used the ICDiamond, Fan RPM Droped from the difference in temp load. Imagine how well this would do if the fan speed had kicked on like it did for the befor temps! HOLY CRAP lol

I would do another test with the fan maxed out , but this MB is junk and cannot alter fan speeds :[

Befor:
tbl.jpg


After:
tal.jpg



4c less idle , 9c less load.

Prime95 Small FFT test (I hope this was the ideal test to run)

Room Temp: around 27c (not 100% accurate)

CureTime: 10mins (So these results may even get better!)

Im very confident that if this was a better Heatsink/cooler the results would have been even more effective. A spring loaded proper mount would have definatly made a better setup.
 
I apologize for the delay. I have been really busy lately but I finally found time to test and post results. Hurray!! :)

Wow, great thermal paste =)
Thanks IC Diamond!

The "before" was using stock thermal paste and stock cooler.
The "after" was using IC Diamond thermal paste and stock cooler.

PCTemps-1.jpg


Coffee

PS: The IC Diamond thermal paste has only cured for about 6 hours during the "after" reading.
 
Here are my results, sorry I am a bit late.

E6600 at stock clocks and voltages on Gigabyte EP45UD3L

All temps are in degress celsius.


IDLE:
Ambient:=======27-------------27
Core 1 Idle:=====37-------------33
Core 2 Idle:=====36-------------31

LOAD: (One hour of Prime95)
Ambient:=======27-------------27
Core 1 Load:=====59-------------56
Core 2 Load:=====59-------------56

Here are graphs contatining the above data and other useful results:


idlerealtemp.png


loadrealtemp.png


Raw data is availble if anyone is interested. :)

Thanks IC Diamond.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom