TechPowerup IC Diamond Results Thread

[t_ski]

Using the system in my sig:

Intel i7 3930K
Asus P9X79 Pro
4 x 8GB Corsair Dominator GT
Triple 7970's (stock coolers)
Triple Asus VW246H LCD's
2 x 120 GB Corsair Force 3 (Raid 0)
500GB WD
Lite-On BluRay
Nickel-plated Heatkiller 3.0 CU
D5 with EK X-top
BIX Quad Rad w/ six Noctua NF-F12 PWM fans on the rad at full speed (4 on one side, 2 on the other to fit around the PSU, which also pulls air through the rad)
BIX 120mm rad w/ one Yate Loon medium speed fan at full speed
Corsair AX1200
Silverstone TJ07

Before

Using MX-4 thermal paste. Temps below were after a 5-pass run of Intel Burn Test.

Ambient temps were 21.3C (measured with my laser thermometer at a random, non-heat-producing item in the room). Loop temps were about 24C, measured at the res.

AFTER

Using ICD24 thermal paste. Temps below were after a 5-pass run of Intel Burn Test.

Ambient temps were 20.4C (measured with my laser thermometer at a random, non-heat-producing item in the room). Loop temps were about 23C, measured at the res.

Note the hole in the IHS is at the top left corner of the paper, just as it is on the chip in the socket.

CONCLUSION

Temps were about the same given the test for ICD24 was with 1C cooler room and loop temp. In some cases, the ICD24 was just a hair warmer.

 

[Kantastic]

Another delay on my motherboard... open box from Newegg apparently means a reconditioned board that's shipped with no packaging whatsoever, just a plain white box (the ones you get back from RMAs) with a label on it.

System still isn't up yet... FML!!!

 

[IC Diamond]

I have been doing these forum survey tests for awhile and thought not much new was to be learned from them then on our last one @ OCUK I had a high% of WB results about 20 -25% as noted earlier. An interesting statistical cluster but I had no end users interested in doing contact impressions.

So now here at TP along with Gilgamesh's images The WB's are from a cursory glance are pretty much all universally poor contact.

On all 20+ surveys we have done we have always had an issue with WB's however the sample size was always small and drew little attention. It will take some time but we are going back through the previous data and filter out the WB tests to combine with the TP results to get a broader understanding.

Normally when we send out the C/P paper we include a self addressed envelope to return for a lab analysis and this is the first time we have not done so and am sorry we did not do it here but we were not anticipating doing it. In any event, if any one here would like a full lab analysis Innovation Cooling will send it out for an independent lab analysis from sensor products for the hard data numbers. Most look to be well under 50% contact and it would be nice to take the guess work out of the results.

Our mailing address is

Innovation Cooling
350 East Main Street
Ansonia CT. 06401
USA

 

[IC Diamond]

Full analysis looks like this and will be emailed to you in an Excel file.

 

[SonDa5]

So now here at TP along with Gilgamesh's images The WB's are from a cursory glance are pretty much all universally poor contact.

That contact paper test was from my i5-3570k with a DT SNIPER water block on a MSI Z68A GD65 G3 MB. I threw away the contact paper so I wont be able to mail it in for full analysis.

OPT33 from Overclock.net stated that I had near perfect contact pattern???


I ordered a EK-LGA115X TRUE Backplate and once I get the new back plate in I am going to see if it helps improve surface area contact with the IHS.


After I do the IHS I am going to delid my 3570k and do a contact pressure on with direct die contact.


IMPORTANT!!!


Is it safe to put IC Diamond directly on the die of a cpu or gpu?

 

[IC Diamond]

That contact paper test was from my i5-3570k with a DT SNIPER water block on a MSI Z68A GD65 G3 MB. I threw away the contact paper so I wont be able to mail it in for full analysis.

OPT33 from Overclock.net stated that I had near perfect contact pattern???


I ordered a EK-LGA115X TRUE Backplate and once I get the new back plate in I am going to see if it helps improve surface area contact with the IHS.


After I do the IHS I am going to delid my 3570k and do a contact pressure on with direct die contact.


IMPORTANT!!!


Is it safe to put IC Diamond directly on the die of a cpu or gpu?

Quite common on notebooks with bare die for several years with no reported issues

Use one of the unused tests and send me that.

Then send me your test results and I will forward to sensor products for analysis. Pressure analysis can be done with an accuracy of about 10% by eye with the previously posted PDF file.

The amount of area in contact is dicey by eye but with their software you can get a pretty accurate number as it calculates the number by individual pixels count so you can subtract the contact area from the IHS area and work a percentage of contact area.

Pressure is done by the pixel density

I have been working with the stuff for years in-house weekly and have done probably a couple of hundred on forums and for individual trouble shooting and even with my practised eye I am probably only 20-30% within a measured result and from an internet Image? My guess is less than 50% contact but I could be wrong they have made some changes to the paper in the last year or two and is more difficult now to make an educated guess on the result.

Best to rely on hard data for contact analysis

What I would like to do is generate charts like the ones below for water cooling

 

[DOM]

Msi HD 7970 ref card stock clocks 925/1375

Before Compound:stock
Ambient Temp:23.8
Idle Temp:37
Load Temp:70

ICD
Ambient Temp:23.3
Idle Temp:35
Load Temp:68

 

[IC Diamond]

That contact paper test was from my i5-3570k with a DT SNIPER water block on a MSI Z68A GD65 G3 MB. I threw away the contact paper so I wont be able to mail it in for full analysis.

OPT33 from Overclock.net stated that I had near perfect contact pattern???


I ordered a EK-LGA115X TRUE Backplate and once I get the new back plate in I am going to see if it helps improve surface area contact with the IHS.


After I do the IHS I am going to delid my 3570k and do a contact pressure on with direct die contact.


IMPORTANT!!!


Is it safe to put IC Diamond directly on the die of a cpu or gpu?

Just thought I would note this as a comparison to your print.

Contact here was 60.35% this was done with the old paper vs new. so not completely contact comparable. Light areas of contact tend to be darker on the new C/P paper, same pixel count but more sensitive so is more easily scanned hence more accurate. While visually more prominent it tends to lead to over estimation when comparing to the old C/P paper which is why I threw out the less than 50% contact estimation. but hard data will give a real number.

The type of contact is just as important as total %, From the image you can see that most of the contact is at the edges with 85 + lbs (at the upper limit of the paper), perhaps showing a cupping of the IHS so that some portion of the heat has to find it way through the IHS to the edges to be removed, obviously not real inefficient.

In a good contact situation you would have a more even distribution. in the example below a max of 97 psi at the edges tapering off to zero towards the centre with an overall force/pressure on the IHS being at only 45 psi.

So for example on your IHS if you were to lap it and had no high spots you would probably have a pressure range of perhaps 50-60 PSI across the IHS with no 85+ PSI high spots and your overall would still be the same force/pressure.

If you remove the IHS and have great contact the decrease in surface area over the IHS will proportionally increase your pressure so you could have something like 90 psi in the contact area even though your overall force might be only 50lbs which is great from a performance standpoint.

With high pressure like that a cautionary note - A slight misalignment on mounting and the die will crack/chip on the edges. I would also make sure the sink is perfectly flat some surface discontinuities could cause a problem, otherwise I have noted good results with the process so the risk may be worth it.

 

[cdawall]

Reserved for my results just got the tube and working on a top overclock to test with. Using Noctua NT-H1 currently...and rather like it so will be interested to see how this compares.

 

[crazyeyesreaper]

updated my post with pressure data IC Diamond 24 paste will have to wait untill new back plate arrives.

 

[AphexDreamer]

Got the tube too. The moment I start using the stuff I'll be back with results. Won't be to long.

 

[cdawall]

Started some of the testing

Rig Specs:

Phenom II X4 B97 3.2ghz@4ghz 1.5625v LLC enabled
Asus CH3F
Corsair H70 with dual 120x38 Nidec Beta V's, 120x25mm shrouds
Room temp of ~23C

Noctua NT-H1 idle testing ~28C

IC Diamond idle testing ~28C

Noctua NT-H1 Prime 95 load testing ~43C

IC Diamond Prime 95 load testing ~44C

Got a couple more to go

EDIT:

Update with Zalman paste

Idle is 31C Load is 44C

 

[IC Diamond]

No C/P images?

 

[cdawall]

No C/P images?

Will do those last after I get the other brands of thermal paste done.

 

[crazyeyesreaper]

got my new backplate in can do the temperature testing soon just waiting for the stuff to cure woot.

 

[BATOFF3]

Here are my results. I didnt allow any cure time.
Overall I am pleased with this tim.
C/P isnt the best result , but Im still happy with the final results.

 

[t_ski]

Do you think you could take a closer picture of the contact paper? It's hard to see between the size and the flash.

 

[BATOFF3]

Added a closer picture of contact paper....best i could get it.

 

[stinger608]

Starting to see some results on here!

Awesome everyone.

 

[stinger608]

Come on folks, lets get some results posted...........

 

[crazyeyesreaper]

if things dont go terrible today ill have my results posted

 

[Kantastic]

Sorry guys, still don't have a working computer. College started and I've been busy. Will send my Asus motherboard for RMA this afternoon since my late class is cancelled. If all goes well, I'll have a system up and running & results in 2 weeks (10 days for RMA).

I still need to send a motherback back to someone on TPU...

 

[3870x2]

Please use the "multi-quote" button or edit your posts instead of posting multiple times in a row.

It is interesting to note that you wasted 4x as much forum space posting the correction.

Knowledge is Power!

 

[t_ski]

It's not necessarily about wasting space as it is considered impolite (spamming).

 

[SonDa5]

Finally got around to remounting my DT Sniper Block again to see if my temps will improve.

Before my contact pressure wasn't parallel with my i5-3570k die.

It was vertical. Here it is.

Now my contact pressure is in a horizontal parallel position with my i5-3570k.

Here is the paper in the socket with IB CPU parallel with contact pressure.

I am hoping that the parallel positioning will lower temps. I will post before and after temps for the new position of mount later on.

I have one more contact paper left which I am going to use for a bare die contact test later on and will also follow up with temp results from running direct die contact with IC Diamond.