OCUK Official IC Diamond/ Perihelion Test Results

[Thekwango]

even with it freshly applied I'm already liking what I'm seeing, gonna let it run a couple or 3 hours tonight and then test tomorrow.

but, if after the IC Diamond has been allowed time to 'cure' temps improve further from what I am seeing, I am gonna be mightly impressed.

I'll also tighten the pressure nut on the venomous x slightly once the ICD has had a couple of hours run time

I know it's early days but this is looking at least the same if not a bigger improvement in load temps from when i moved from a hyper212+ to a venomous x..

will update with a few screenies and results in the next day or so

 

[El_Commi]

System:

CPU AMD Athalon 64x2 4400+ Brisbane @ 2300MHz
Motherboard: An-m2HD (MC P68)
RAM: Corsair XMS 4GB@ 800MHZ
Stock AM2 cooler, with 2 year old paste I believe to be Akasa AK-455-5G

Tested using LinX0.6.1, with 2048 MB memory used, with 64bit mode enabled @ normal priority

class going 10 runs.

Ambient temps: 16.9
Idle: Core 1: 21-22^c Core 2: 27^c http://img231.imageshack.us/img231/752/amd6400tempsidle.png
Load: Core 1: 47^c Core 2: 56^c

http://img861.imageshack.us/img861/3853/amdx6400loadstock.png

Stopped at 5 runs as there was no noticible difference in high temps.
Allowed several hours for solution to cure.

IC Perihelion

Ambient: 18.9^c

Idle: Core 1: 22^c Core 2: 27-28^c http://img844.imageshack.us/img844/2605/perihelidle.png
Load: Core 1: 48-40^c Core 2: 57-59^c

http://img151.imageshack.us/img151/3413/perihelionload.png





System:

CPU: i5 2500k @ 45Ghz
Motherboard: Asus P8p67 pro
Ram: g.skill clocked @ 1333MHz
Cooler: Arctic freezer pro 7 rev 2 using mx4

Testing using LinX 0.6.4, using 2048 MB memory, 64bit more enabled normal priority class,

going 10 runs.

Mx-4
Ambient temps: 23.5^c
Idle: Core 1: 36^c, Core 2: 40^c, Core 3: 33^c, Core 4: 36^c

http://img834.imageshack.us/img834/1886/mx4idle.jpg
Load: Core 1: 73^c, Core 2: 79^c, Core 3: 77^c, Core 4:77^c

76^chttp://img24.imageshack.us/img24/3609/mx4load.jpg

IC-Diamond 24
Ambient temps: 24.7^c
Idle temps: Core 1:35-36^c Core 2:39-40 Core 3: 33^c Core 4: 37-38^c
http://img9.imageshack.us/img9/240/icdiaidle.jpg
Load temps: Core 1: 74^c Core 2: 80^c Core 3: 79^c Core 4: 77^c
http://img51.imageshack.us/img51/8840/icdiamondloadnocure.jpg

Allowing computer to run for approx 3 hours to allow the solution to cure, before running

tests again.

IC-Diamond 24 after 3 hours curing time and tightening of Heatsink screws

Ambient temps: 26^c
Idle Temps: core 1: 37-38^c, core 2: 40-41^c, Core 3: 33-34^c, Core 4: 37-38
http://img696.imageshack.us/img696/7913/icdiamondcuredidle.jpg
Load Temps: Core 1: 74^c, Core 2: 80^c, Core 3: 79^c, Core 4:78^c
http://img854.imageshack.us/img854/559/icdcuredload.jpg

Seems to be running about 1-2^c higher than my Mx-4.

 

[IC Diamond]

mx4 ICD 1 ICD2
73 74 74
79 80 80
77 79 79
77 77 78

Avgerage 76.5 77.5 77.75

Ambient 23.5 24.7 26

Temp rise above ambient 53 52.8 51.75

Have to subtract ambient so ICD - 0.2 C first run , -1.25 C 2nd run

 

[El_Commi]

Is that within acceptable ranges, do you think?

I have a feeling I may have underused the compound. I applied a blop dead center of the cpu, about the size of a garden pea. But after a little reading seems I should have used a larger pea

I'll give it a blast when I'm home after work this evening, so we can compare. Haven't tweak my volts yet, so results should be still admissible.

 

[bulldog147]

IC Diamond 24 is just a waste of money ! MX4 is the best thermal paste and I can see here it not much the different between MX4 and ICD 24. So, paying an extra fiver for ICD 24 is just silly to be honest !

 

[bulldog147]

I believe manufacturers of both CPUs and Heatsinks should pay more attention to the importance of having flat surface and to facilitate large contact area.

I do agree with you +1

 

[IC Diamond]

IC Diamond 24 is just a waste of money ! MX4 is the best thermal paste and I can see here it not much the different between MX4 and ICD 24. So, paying an extra fiver for ICD 24 is just silly to be honest !

So let's put it to the test with a shoot out between the two

I'll put up the IC Diamond and if I can come to a reasonable arrangement with OCUK for a couple hundred tubes I'll pay for the MX-4 - we can test fro both performance and reliability.

I have been thinking of changing the format and going one to one from here on out this would be a great test if we can get enough people to participate..

I will even add a contact and pressure test kit with each sample, lab analysis is to expensive but Raw prints will reveal a lot and be comparable to IC library of image analysis to date

 

[IC Diamond]

I do agree with you +1

Personally I think the whole bowed base concept is to edgy "if" it is a viable tweak to hit an optimum point you are relying on too many uncontrollable variables with IHS contour and hardware mounting inconsistencies.

With the two lapped examples we a 5C and a 7C improvement - the first high end sink manufacturer that incorporated it would kill the competition.

I think it is most likely a manufacturing issue as noted earlier - the base is warped on heat assembly and the final assembly is too awkward to machine flat.

 

[El_Commi]

After reapplication of paste- With no curing time allowed.

Ambient temps: 24.5^c
Idle: Core 1: 33-34^c, Core 2: 37^c, Core 3 30^c, Core 4: 34^c
Load: Core 1: 70^c, Core 2: 74-75^c, Core 3: 73^c, Core 4: 72^c

Overall, a much much more impressive result. More along the lines of what I was expecting. IT took much longer to reach max temps, and when it did- they rarely stayed there long. Usually dropping back down 1-2^c.

I'll not mess with my volts/clocks tonight in case Ic-diamond wants more tests tomorrow

 

[IC Diamond]

How old was the original install of MX4?

 

[DarkBahamut]

Applied my ICD yesterday and gave it some time to bed in and did some testing today. Some nice gains all round and a good improvement on my GPU even with completely stock cooling still in place

Core i7 2600k O/C 3.8GHz @ 1.08vcore (1.6GHz @ 0.800vcore idle) - NH-D14 heatsink/fan on silent/slowest fan speed (800RPM). Antec P183 4 case fans all on 700-800RPM (silent mode). Prime 95 1hour large FFTs.

Arctic Silver 5:
Ambient: 22.0C
Idle: 27.75C (28, 25, 29, 29)
Load: 51.5C (51, 50, 53, 52)

ICD:
Ambient: 22.0C
Idle: 26.5C (26, 24, 28, 28)
Load: 49.5C (49, 47, 52, 50)



Radeon 6970 stock cooling. 915/1430 stock voltage. Intensive GPGPU work load for 1 hour (more demanding and stable than game testing)

Stock Compound:
Ambient: 22.0C
Idle @ 25% manual fan speed: 41C
Load @ 60% manual fan speed: 89C

ICD:
Ambient: 22.0C(idle testing) 23.5C(load testing)
Idle @ 25% manual fan speed: 38C
Load @ 60% manual fan speed: 80C

 

[El_Commi]

How old was the original install of MX4?

Not much more than a month on the i5. 6 weeks tops.

*edit*
Have screenshots of the second round of testing, forgot to upload them this morning. Will do so later when I'm home from work.

*ninja edit* Screenies as promised:
http://img842.imageshack.us/img842/5626/reapplicationidle.jpg
http://img39.imageshack.us/img39/7173/reapplicationload.jpg

 

[Liquid]

After what seems an eternity, I can finally post my results. Had a hell of a time and three motherboards later lol.

System - 17-2600k running at 4.8ghz on a Noctua D14 cpu cooler.

All tests were taken applying compounds the night before, running for a couple of hours to bed in and then taking measurements the day after, times did vary because I was making sure the ambient temperature was the same each time.

Used prime for full load measurements. All tests were done running prime for 1 hr, and 3 times, with the average being noted. I also averaged the 4 cores to simplify the table.

Of course, some variance may have occured with the amount I put on. I must admit, I would have thought the mx-2 would have done better. As far as I am concerned the IC Diamond is well worth the money and certainly what I will be using. I am also going to do my video card when I sort a cooler out too.

 

[bulldog147]

That's pretty good. Did you put small sized pea in the center on the cpu chip and attached N14 heatsink and let it spread itself ? Did u use Noctua D14 stock fans 140mm and 120mm ?

I had a Noctua D14 on my other pc i7 920 and might put Noctua D14 onto my new i7 2600K but I probably wait until I order some IC Diamond 24.

 

[shadow_boxer]

Right so about time I tested this too. Received tubes quite a while ago. Replacing my Titan Fenrir with the Antec Kuhler H2O 620.

Tested with Arctic Silver 5 now for IC Diamond.

What is the best way to apply IC Diamond? With AS5 I applied small pea sized amount to CPU and spread with plastic credit card.

I used it on my HTPC but found that using the spread method was difficult because of the thick viscosity of the paste and abandoned it and went for AS5.

 

[IC Diamond]

Right so about time I tested this too. Received tubes quite a while ago. Replacing my Titan Fenrir with the Antec Kuhler H2O 620.

Tested with Arctic Silver 5 now for IC Diamond.

What is the best way to apply IC Diamond? With AS5 I applied small pea sized amount to CPU and spread with plastic credit card.

I used it on my HTPC but found that using the spread method was difficult because of the thick viscosity of the paste and abandoned it and went for AS5.

Application Notes -Amount of Applied Compound - FAQ
*
*
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 30mm X 30mm 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 0.15 ml which approx. equals our 5-5.5mm size bead on center.

IF YOU ARE NOT SURE PEEL BACK THE LABEL TO SEE ML REGISTRATION MARKS AND USE 0.15-2ml EXCESS JUST SQUISHES OUT - I HAVE MEASURED THIS WITH A MICROMETER @ 50 PSI AND AT ROOM TEMP YOU REACH 95% OF THE BOND-LINE THICKNESS (BLT) IN TWO HOURS 99% OR MAYBE 1/2 DEGREE OVERNIGHT AFTER THE INITIAL 2 HOURS
*
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 AKA cusp formation and viscous fingering.

Fast Simulation of Laplacian Growth

http://www.youtube.com/embed/AHmgPtOJsYk
Laplacian Growth in thermal compounds

Laplacian Growth - a good visual explanation

http://www.physics.ubc.ca/pitp/archives/theory/2004talks/wiegmann.pdf


http://innovationcooling.com/applicationinstructions.htm

 

[Liquid]

That's pretty good. Did you put small sized pea in the center on the cpu chip and attached N14 heatsink and let it spread itself ? Did u use Noctua D14 stock fans 140mm and 120mm ?

I had a Noctua D14 on my other pc i7 920 and might put Noctua D14 onto my new i7 2600K but I probably wait until I order some IC Diamond 24.

Yep, as advised I put a pea sized blob dead in the centre then let the cooler spread it out under attached pressure. I am using the standard fans that come with it at full speed they offer. I love that D14 and IC Diamond seems the perfect companian for it.

 

[Mr Krugga]

Intel Core i5 750 VID 1.207v, @ 3.8GHz 1.28v BIOS, 1.29v idle, 1.33v load
Zalman CNPS10X Extreme + Zalman STG-2 and IC Diamond 7


Curing time:

Zalman STG-2 - months
IC Diamond 7 - over 2-3 hours of stress testing


Zalman STG-2 results:

IBT v2.51 Custom 2560MB - 20 runs:

Ambient Temp: 21c
Idle Temp: 32-35c
Load Temp: 69-77c

Prime95 blend over 1h - 1h 30min:

Ambient Temp: 22c
Idle Temp: 31-34c
Load Temp: 68-76c

IC Diamond 7 results:

Prime95 blend over 2h:

Ambient Temp: 23c
Idle Temp: 31-34c
Load Temp: 62-65c

IBT v2.51 Custom 2560MB - 20 runs:

Ambient Temp: 22c
Idle Temp: 31-34c
Load Temp: 64-67c

Idle Temp with C-States/C1E/ESET: 25c
Ambient: 21c

• Possibly another update with Phenom II X4 B55 with CoolerMaster Hyper 212+
• Note: all Power Saving options are turned off, with them on I will drop the idle temps by 5c and possibly the load temps by 1c

 

[varkanoid]

PLEASE READ THOROUGHLY BEFORE TESTING/POSTING YOUR IC DIAMOND OR PERIHELION TEST RESULTS!!!


.....

I am intended to test ICD and ICP on my GTX480 only but what GPU tests should I run and what is the best way of application ? It doesn't seem to mention on your website and the video card tests link goes to a page not found.

 

[Exsurgo]

i7 920 @ 4.2Ghz with 1.34 volts and 1.32VTT
EK Supreme HF Copper
2 x Thermochill PA 120.3s

Original compound was MX-2
All tests were 15min of LinX 4900mb memory used

** Application method was the cross method (see below) **

Before Compound:
Ambient Temp: 19
Idle Temp: 31
Load Temp: 64

ICD 24
Ambient Temp: 20
Idle Temp: 33
Load Temp: 63.5

During my first test I saw that the IC diamond 24 was >1c higher than the MX-2 under load even after > 8 hours cure time. I realised that this could well be because my MX-2 application method had been the cross approach. I reapplied the Diamond in the cross pattern and got the results above.

The diamond is a very thick paste that is tricky to apply to the IHS. This could be the reason why the pea method produced weaker results. MX-2 spreads easier from my experiance.


**EDIT**

The difficulty in spreading with the pea method with IC Diamond 24 could be to do with the EK Supreme's poor mounting mechanism