TechPowerup IC Diamond Results Thread

[fullinfusion]

... bump for a replie to Sonda5's question..

Guess I'll answer my own question then

It must be the same shit as last year... Hmmm

 

[IC Diamond]

ICD24 same as ICD7 - Numbers denote Carat weight - 1 Carat = 0.2 gm

Updated library of contact images on post 116 on the previous page to include the EK Supremacy Full Nickel. Not looking good for the WB's with the exception of one of the Koolance tests

EK Supremacy FULL NICKEL courtesy of Gilgamesh

 

[adulaamin]

adulaamin -Did not get the type/ manufacturer of your sink

Oh sorry... It's a Cooler Master Hyper 212 Evo

I did the test twice coz I thought I did it wrong the first time... Same result... I might try to lap the base although I've never done it before that's why I'm kind of hesitating hahaha...

 

[cdawall]

Been giving the ICD a good good burn in time with the phenom/heavy watercooling setup. Stuff is doing pretty well but almost no improvement next to the Noctua from before. Will post real world number shortly. So far it is a 40C load temp either way.

 

[1freedude]

Ok, SABERTOOTH x58 i7 920 at stock cmos:

Idle at desktop: 35
Intel Stress Test from Extreme Tuning Utility: 43
F@H 8 threads: 56

This is using 4GB 1333 DDR3 Triple Channel, and an old application of IC Diamond.

I will remove Dell heatsink, clean, and check with C/P.

Ambient: 70.2 F

With fresh paste, F@H temp raised 1-2.

C/P shows really weak results. Not suprised, as the pressure is delivered via springs. Later (not tonight), I will mod it to use nuts and bolts.

Time to finish Neuromancer.

 

[TRWOV]

TRWOV's results

TEST SYSTEM:
CPU: Intel Core 2 Extreme QX6800 B3 @ 2.92Ghz 1.5v (stock), lapped
MoBo: Asrock Conroe865PE
Cooler: Cooler Master Hyper N520

Tested compounds:

- Cooler Master standard thermal compound. No cure time.
- Artic Cooling MX-4. No cure time.
- IC Diamond 24 carat. ~10 hours of cure time, 5 cold-hot cycles.


Pressure pads results:

Test 1: Heatsink stock
Test 2: Heatsink slighty lapped
Test 3: more lapping




Results:

Intel Burn Test:

Cooler Master standard thermal compound (no cure time, TEST3 pattern):
Max avg (4 cores): 75C
Min avg (4 cores): 39C

IC Diamond (+6 months of cure time, TEST1 pattern, low pressure):
Max avg (4 cores): 68.5C
Min avg (4 cores): 40.75C

Arctic Cooling MX-4 (no cure time, TEST3 pattern):
Max avg (4 cores): 68.5C
Min avg (4 cores): 37.75C

IC Diamond (~10 hours of cure time, TEST3 pattern, high pressure):
Max avg (4 cores): 67.25C
Min avg (4 cores): 38.5C

Pics:

Spoiler

Cooler Master standard thermal compound

IC Diamond (TEST1 pattern, low pressure)

Artic Cooling MX-4

IC Diamond (TEST3 pattern)

OCCT:

Core 0 is the hottest so I'll use that graph only. I should have left HWmonitor running along with OCCT

Cooler Master standard thermal compound (no cure time, TEST3 pattern):

IC Diamond (+6 months of cure time, TEST1 pattern, low pressure):

Arctic Cooling MX-4 (no cure time, TEST3 pattern):

IC Diamond (~10 hours of cure time, TEST3 pattern):

Conclusion:

It seems that the IC Diamond compound needs heat and good contact between the heatsink and CPU to truly shine as its performance increased with improved pressure and although idle temps are nothing to write home about (the standard CM compound comes close) at high temps the performance is superb. Also this series of test just opened my eyes on how bad contact some heatsinks provide, I think I'll buy some pressure paper in the future.

I have to comment on MX-4, though: with no cure time it performed within a 1C range in relation to IC Diamond. Still, IC Diamond performed just as well even with substandard mounting pressure (TEST1 vs TEST3).

 

[TRWOV]

and just now I learn that IC7 and IC24 are the same thing. The 7 carat and 24 carat come from the weight changing post accordingly.

 

[SonDa5]

and just now I learn that IC7 and IC24 are the same thing. The 7 carat and 24 carat come from the weight changing post accordingly.

I knew they were the same. ICD 24 has been out for over a year as well. I used IC24 over a year ago and it has worked well. I thought this free sample was to get a totally new formula but in turns out its just to get more testing done. It's all good.

 

[EarthDog]

I posted in the 'request thread'.. oops! Anyway, here are the results...again:

Ok...... the eagle landed last week and went up H2H with MX-4....

Environment was 21C under custom water for the tests. I remounted the Diamond a couple times to make sure i was getting a good spread/mount as this stuff is STICKY and doesnt spread well it seems. That said after a good contact I saw these results:

Stock:
MX-4 idle - 27C (average across cores)
IC D idle - 27C ( " " " )
MX-4 load - 54C ( " " " )
ICD load - 53C ( " " " )

Overclocked - 4.9Ghz 1.5v - (no power saving on constant v and clocks)
MX4 idle - 30C
ICD idle - 29C
MX4 load - 93C
ICD load - 91C


Ugh, the viscosity and the ability of this stuff to spread is thick, wow. It doesnt even like to spread with a credit card or anything it just like clumps and rolls off the processor. Gotta do something about that if possible with the materials you use!

Thanks for the sample, solid product guys.

 

[crazyeyesreaper]

lol in the regular thread the guy pretty much says dont spread it use pressure only

 

[EarthDog]

Right... after multiple failures with getting a good spread (tried increasing quantities of product, mounting the block (Apogee XT) in different ways), it just wouldnt spread well. So I tried to spread it with a CC... that failed even worse. It turns out that copious amounts of this product (literally pea sized, not rice - as the instructions stated) were needed with my setup to achieve adequate coverage.

I dont have this issue with MX-2/4, AS5 and the temperature difference wasnt that great.

Single case though, I would imagine the problem is with the mounting pressure being limited on the Apogee XT setup vs others that are not? No clue.

 

[crazyeyesreaper]

yea mounting pressure etc matters, i mean hell i saw an 8c drop in temps at the SAME ambient room temps on my Ivy bridge system compared to MX4, i dont know why i did multiple tests of each but temps stayed roughly the same within 1c on each test, then again compared to many in this thread i have one of the best mountings in terms of surface contact area and pressure and both together make all the difference it seems.

 

[EarthDog]

Yeah.. temps were fine with any product, mounting pressure was good and as complete as I have seen it (sorry, no photos of that... ) so it wasnt that.

 

[IC Diamond]

EarthDog

5.5MM Bead id recommended - compression spread should work fine unless you have extremely light pressure then if that's the case some effort should be made to shore up the mount to provide more pressure as you are likely in the 30lb range and would gain something like 5C even on a generic paste.

But as your results are more or less in the expected range pressure probably is fine, a little tweaking perhaps but not bad.

Alternatively you can mix a drop or two of solvent like acetone or IP to thin it to your desired consistency then let it dry Completely about 10 min. then apply sink.

Do make sure it is dry otherwise it will leave voids in the joint.

 

[EarthDog]

Thanks IC D. I tried tightening down two screws at a time, one at time in a circle, and crossed pattern, it just didnt do well until I literally had a pea sized amount on there (like your instructions state).

My concern isnt so much the spread as it worked out when I followed the instructions, but the amount of applications /tube /$ vs other products.

 

[IC Diamond]

Thanks IC D. I tried tightening down two screws at a time, one at time in a circle, and crossed pattern, it just didnt do well until I literally had a pea sized amount on there (like your instructions state).

My concern isnt so much the spread as it worked out when I followed the instructions, but the amount of applications /tube /$ vs other products.

Does not matter - Paste application is more of a volume fill + area coverage. Better to have enough compound on the mount for optimal performance than to have to redo your mount which wastes more compound on a reapplication.

This is why some people have poor results with the older metal pads they were so thin there was not enough material to bridge the void gaps on IHS?sinks with light contact

On the 24 carat syringes sample you received should be enough for 10 applications

Application Notes -Amount of Applied Compound
*
*
There is always some debate on amount of applied compound to use, our approach is based on a best to worst case scenario with a one size fits all so the user can experience best odds of a successful mount out of the gate without having to redo his mount.
*
*
Say that Intel has a spec of Flat and parallel // to .002. - At the extreme end you get something like the 2 images of the candidate below, to fill that gap/volume you need .05 ml volume of compound for a 30mmk X30mm area about the size of a grain of rice.
*

*
Our difference with the rice method comes from the other side of the equation The heat sink base, while many HS bases are good some can be off as much as the IHS and to cover the contingency of irregular contour on both sink & IHS and as few people lap to correct the problem (5%?)and fewer lap to perfection so we add another .05 ml for a total of 0.1ml.
*
Another factor associated with irregular contours is another mil or two layer of compound is added on the high spots depending on the pressure applied. To fill that gap we add another .05 ml for a total recommended amount of 1.5 ml which approx. equals our 5-5.5mm size bead on center.
*
Thermal compound is a volume application to fill gaps, not one of weight, liquid spread-ability etc. and our opinion is that whatever the compound you still need the equivalent volume to close that gap.
*

*
Additionally it is important to apply enough compound, because as the paste thins it spreads it reaches a maximum zone where air is reformed into the paste application between the sink and his leading to decreased performance and early compound failure. This effect is well-documented and known as Laplacian Growth or cusp formation and viscous fingering.

Fast Simulation of Laplacian Growth

YouTube - Broadcast Yourself
Laplacian Growth in thermal compounds

 

[TRWOV]

I pulled my block off yesterday for some other work. I found that there was a slight stain in the IHS on the CPU. The waterblock did not show anything - at least, it was very clean in the contact area, but the outer edge was slightly tarnished. I'll post pics up for your review later.

Use some lemon juice to clean the stains.

 

[NAVI_Z]

I've been reading this thread quite closely and was wonderin if anyone out there had

Indigo Extreme Engineered TIM to compare to IC-D 24? The pressure test with the paper is

quite an eye opener.I was also under the impression that the CPU had more of a "concave"

than the HS.I say this because I've read many reviews on HS's and a some water blocks

made by some manufactures that make their blocks with a slight "concave" for "optimal"

contact surface area n such.

 

[IC Diamond]

I've been reading this thread quite closely and was wonderin if anyone out there had

Indigo Extreme Engineered TIM to compare to IC-D 24? The pressure test with the paper is

quite an eye opener.I was also under the impression that the CPU had more of a "concave"

than the HS.I say this because I've read many reviews on HS's and a some water blocks

made by some manufactures that make their blocks with a slight "concave" for "optimal"

contact surface area n such.

LM are great thermal performers in fact we entertained the idea of a LM product before we settled on developing our own compound. I would say that LM's are roughly equivalent to a solder joint but our analysis was that they are less forgiving than IC Diamond in poor contact situations but are much better in light pressure mounts as it hits it's BLT with little resistance. In some internet reviews you will see IC Diamond performance a degree or two better than LM's this is due to "Real World" contact issues as I would expect them to be slightly better although a degree or two of test error should be factored in on any individual test. I had one test LM comparison on a German giveaway where ICD was applied for a 20C improvement, pretty much a contact issue I am sure.



So inconsistent end user results is one issue another is cost per application is quite high also in our opinion they are not real user friendly on application or clean up.

LM's despite being on the market for 6 or 7 years are approximately 0.25% of the market if our overall forum sample giveaway's of approx 850 end user test comparisons is any indicator of user preference although it could be skewed due to extra effort in removal and application or perhaps they are happy with their result and see no need to try any thing else.

In any event for whatever reason your average Joe overclocker selects something other for the time being and we believe our initial marketing analysis was correct.

It is hard to separate design strategies from internet urban myth. Intel spec on the IHS is Flat and // to within .002 and from the contact patterns they have a broad range of profiles from concave to convex and or some combination of the two and I see no Intel spec that spec's a concave IHS.....

Concave makes no sense to me in that if you have a concave pattern profile w/bowed or convex sink along with a sink that is larger than the IHS the sink ends up resting on the edges of the concave (IHS) with little or no centre contact which seem to match those patterns like the one below

All in all a flat sink would cover(contact area) more irregular IHS profiles than one that is biased in a particular way. The only argument for it I can think of is that with the reduced contact area where you do have contact the pressure is increased, there may be other alternative reasons that are not obvious to me such as design/manufacturing tradeoffs

 

[KainXS]

k got mine today and tested on my c2q and obviously its lapping time

my old baby

Core 2 Quad Q8400@3.5Ghz(Downclocked from 4)
Asus GTX 460
Artic Cooling Geminii+1 120x38mm + 4 Case fans
Antec 300

I still love/d me some AS5

OCCT

before IC Diamond(This is the best I could ever get with AS5 on my PC)
Ambient is 28C
Idle is 38C
Fully cured

IC Diamond
Ambient is 28C
Idle is 38C
has cured for about 3 hours now

over all the cores I am getting a solid 3C to 5C drop, very cool

GTX460 - Direct CU

Kombuster

Before(With AS5)
Ambient is 28C
Idle is 39C
Fully cured

IC Diamond
Ambient is 28C
Idle is 32C
3 hours cure time

all I can say is damn, AS5 got its ass kicked

but I need to lap with that pressure test what kind of lap would u recommend IC Diamond

 

[johnspack]

Hmm, I should get in on this... oced first gen i7 can get all the help it can!

 

[NAVI_Z]

LM are great thermal performers in fact we entertained the idea of a LM product before we settled on developing our own compound. I would say that LM's are roughly equivalent to a solder joint but our analysis was that they are less forgiving than IC Diamond in poor contact situations but are much better in light contact mounts as it hits it's BLT with little resistance. In some internet reviews you will see IC Diamond performance a degree or two better than LM's this is due to "Real World" contact issues as I would expect them to be slightly better although a degree or two of test error should be factored in on any individual test. I had one test LM comparison on a German giveaway where ICD was applied for a 20C improvement, pretty much a contact issue I am sure.



So inconsistent end user results is one issue another is cost per application is quite high also in our opinion they are not real user friendly on application or clean up.

LM's despite being on the market for 6 or 7 years are approximately 0.25% of the market if our overall forum sample giveaway's of approx 850 end user test comparisons is any indicator of user preference although it could be skewed due to extra effort in removal and application or perhaps they are happy with their result and see no need to try any thing else.

In any event for whatever reason your average Joe overclocker selects something other for the time being and we believe our initial marketing analysis was correct.

http://innovationcooling.com/Sept30condensedmultiforum.PNG



It is hard to separate design strategies from internet urban myth. Intel spec on the IHS is Flat and // to within .002 and from the contact patterns they have a broad range of profiles from concave to convex and or some combination of the two and I see no Intel spec that spec's a concave IHS.....

Concave makes no sense to me in that if you have a concave pattern profile w/bowed or convex sink along with a sink that is larger than the IHS the sink ends up resting on the edges of the concave (IHS) with little or no centre contact which seem to match those patterns like the one below

http://sphotos-a.xx.fbcdn.net/hphot...403x403/225872_539263852757721_82929601_n.jpg


All in all a flat sink would cover(contact area) more irregular IHS profiles than one that is biased in a particular way. The only argument for it I can think of is that with the reduced contact area where you do have contact the pressure is increased, there may be other alternative reasons that are not obvious to me such as design/manufacturing tradeoffs

thanx for such an in depth response.

 

[itsakjt]

So here is my list of tested products
AMD Phenom II X4 with stock cooler at 3.7 GHz 1.375V and IMC at 2.8 GHz 1.275V
Before Compound:Cooler Master Thermal Fusion 400
Ambient Temp:28-29 degree C
Idle Temp:34 degree C
Load Temp:61 degree C

ICD
Ambient Temp:30 degree C
Idle Temp:34 degree C
Load Temp:58 degree C(3 degrees difference even with ambient slightly more.)

HP 430 laptop(Pentium Dual Core 2.2 GHz Sandy Bridge)
Before Compound:Cooler Master Thermal Fusion 400
Ambient Temp:30 degree C
Idle Temp:40-42 degree C
Load Temp:79 degree C

ICD
Ambient Temp:30 degree C
Idle Temp:40-42 degree C
Load Temp:69 degree C(Whoa that is a 10 degree difference)

Sapphire Radeon HD 6770 1 GB DDR5 copper cooler 960 MHz core and 1290 MHz mem stock voltages

Before Compound:Cooler Master thermal fusion 400
Ambient Temp:30 degree C
Idle Temp:42 degree C
Load Temp:81 degree C

ICD
Ambient Temp:30 degree C
Idle Temp:42 degree C
Load Temp:76 degree C(5 degree C difference)

So this thermal compound totally rocks. Thanks for the awesome giveaway.
The key to this being awesome is the thickness. It performs awesome in cases where there is extremely high pressure like that of my laptop or GPU cooler. Also the CPU temp was reduced significantly. Thanks to you all again. I have posted some pictures and screenshots on the giveaway thread and in this one too. This is just summarized.
Stay tuned for a pressure test also and for another Phenom II X4 955 that is now maxing at 66 degree C(overheating).

 

[Novulux]

Stock cooler and paste(i3-2100)
Ambient: 25.6
Idle: 37
Load: 65

ICD
Ambient: 27.7
Idle: 37
Load: 64

Gonna try reapplying later, but I just wanted to get my post in before 30 days..

 

[SonDa5]

New contact pressure test with DT SNIPER water block directly on IB i5-3570k.