[JoeC]

User- Comparison paste-Result Red AS5
EvilGenius * AS5 -1.83
Spamz0r * AS5 -6
CD_TOASTER* AS5 -2
t8y* Stock+.5
Hennels* AS5+0.5
Butcher9_9* AS5 -2
Crash Dummy * Stock paste -3
OsiC* Stock paste -1
stuii*AS5-1
bees_man*AS5+.05
brendohall*AS5+3.25
cbjaust*AS5+8.4
jimmy*AS5-.5
maev*AS5-2
Aetherone*AS5-6
t8y*AS5-.75
terrastrife*AS5-10

User- Comparison paste-Result yellow MX2
EC MEISTER *MX-2 -2
buzzen * MX-2 -2.25
Lardman
CPU -1
GPU1-6.1
GPU1+.5

User- Comparison paste-Result Blue Stock and miscellaneous
redw0lf * Stock-10.25
boogerthe2nd * Thermalright-7
swiftyb * Stock paste-4
tfroggydawg *noctua paste-3.25
JohnZ*Stock Paste-6.1
vchoy* Stock paste -11
Boomsie*Stock-0
boogerthe2nd*stock_6
K3nii*ASC*-2
AussieTemplar*TX2-1.5
jimmy*stock-8.5
NiSlo*Stock=.01
Aetherone*stockWG-10
AetheroneStockintel-14

Updated results OCAU

[JoeC]

I added OCAU to the overall to see what it would look like. With the 160 user tests graphics are getting a little scrunched I will have to figure another way to format it.

[JoeC]

An ICD7 review from the Greeks

[Impar]

Greetings!

XbitLabs tested in CPU and GPU:
http://www.xbitlabs.com/articles/coo...e-roundup.html

[Omsion]

Originally Posted by Impar

Greetings!

XbitLabs tested in CPU and GPU:
http://www.xbitlabs.com/articles/coo...e-roundup.html

Competition is certainly high in the top tiers.
I wonder if you guys have enough samples from people using some of the newer high performing paste to draw any concrete conclusions - I think it's fairly determined that the older generation of pastes have had their performance superseded.

[JoeC]

Originally Posted by Omsion

Competition is certainly high in the top tiers.
I wonder if you guys have enough samples from people using some of the newer high performing paste to draw any concrete conclusions - I think it's fairly determined that the older generation of pastes have had their performance superseded.

I agree, the better pastes that have come out in the last year are used so rarely it is hard to get a large enough sample from users or the review sites. Against AS5 and the stock pastes the trends are quite clear, the better competitors while the samples are low they are growing and trending well for ICD but I would like to see at minimum 20-30 I only have a max of around 10 for a couple of them and only 4-6 for for the others with 450 sample sent out so far, but even with the small sample size the trends is favorable for ICD.

The web site review circuit for pastes is really limiting in that during the last year only a handful, paste roundups were done due the extensive work involved unlike heat sinks which are reviewed daily. Hard to do averaging mental or otherwise with such a small sample which is one reason we do the give away/beta tests.


I would expect that 1 or 2 of the better pastes will group pretty close to ICD - We are at the point of diminishing returns, in our testing the temp drop on a one inch copper test die was about 1 C @ 100W between the die and sink, even doubling the performance would only net 1/2 C and theoretical perfect joint 1 C. Not much headroom for improvement...... The important factors that will start to emerge will be quality and reliability which is where ICD has significant advantages... but what do I know? I am biased hence the program for user beta testing, people can judge for themselves


These are the large roundup results to date it took a year to get to get 5

Xbitlabs



FRAZPL



HardwareLogic



Greece



BenchmarkReviews

[JoeC]

French Review

[trents]

I tried ICD7 recently as it came free with the TRUE I won on an ebay auction. It seemed to give about the same temps as AS5. What I didn't like about ICD7 was its extremely high viscosity. It was like trying to spread road tar or bondo that was almost set up. If you use it you have no choice but to apply a pea-sized dollop on the cpu as the manufacturer web site instructs you to do and then clamp the cooler base to the cpu. You really can't smear it around to make sure you get complete coverage of the mating surfaces like you can with AS5. When I removed the cooler a few days after trying the ICD7 I could see the compression only gave me about 75-80% coverage and it was uneven. I don't know, maybe the free ICD7 that came with the TRUE was used and dried up or something. I went back to AS5.

[JoeC]

Originally Posted by trents

I tried ICD7 recently as it came free with the TRUE I won on an ebay auction. It seemed to give about the same temps as AS5. What I didn't like about ICD7 was its extremely high viscosity. It was like trying to spread road tar or bondo that was almost set up. If you use it you have no choice but to apply a pea-sized dollop on the cpu as the manufacturer web site instructs you to do and then clamp the cooler base to the cpu. You really can't smear it around to make sure you get complete coverage of the mating surfaces like you can with AS5. When I removed the cooler a few days after trying the ICD7 I could see the compression only gave me about 75-80% coverage and it was uneven. I don't know, maybe the free ICD7 that came with the TRUE was used and dried up or something. I went back to AS5.

This is one of the more typical/common troubleshooting forum requests - "I bought "such and such" sink/paste and am not seeing the expected result" the universal replies in the first 5 posts are "check your contact" and "is the sink on tight?"

Our contact and pressure testing validate the forum users recommendations. 80% see a result out of the gate, the 20% that did not were indicated to be poor contact and/or poor pressure

Reference - user contact and pressure testing


The unevenness of your applied paste suggests it could be one of those 2.

Inadequate pressure and the paste will not spread evenly. If you read the threads on ICD7 testing the TRUE performance was spotty But was improved by those that reworked the mounting hardware with shims or bracket mods.

Poor mating contact will evidence the same unevenness. If you have not lapped the IHS or sink that could be the problem.

Hope this helps

[trents]

"Inadequate pressure and the paste will not spread evenly. If you read the threads on ICD7 testing the TRUE performance was spotty But was improved by those that reworked the mounting hardware with shims or bracket mods."

Actually, I did not try ICD7 on a TRUE. At the time, I was using a Zalman 9700LED which mounts on an AM2 with a spring clip. But if TRUE users are having spotty results with ICD7 due to inadequate contact that concerns me as the TRUE has one of the best, high pressure mounting systems I know of right out of the box, at least for the AM2 application. That's what concerns me with the high viscosity of the ICD7. It requires an unusual amount of pressure in order to make it spread adequately. People shouldn't be having to make mods to the cooler mounting system to make it work properly with a certain brand of thermal grease.

[JoeC]

That's what concerns me with the high viscosity of the ICD7. It requires an unusual amount of pressure in order to make it spread adequately. People shouldn't be having to make mods to the cooler mounting system to make it work properly with a certain brand of thermal grease.

The heat sink manufacturers should be doing a better job. It's not the compound it's the mounting hardware or in some cases operator error

All compounds are subject to a pressure performance curve as noted in the chart below - other pastes suffer to the same degree - It's nothing unusual at all - 80% see a result right out of the gate with no mods at all and the 80% of those one's that do not were determined to be a pressure or contact issue. hence my recommendation. Any single test is anecdotal which is one reason to do the beta forum testing. See attached chart

Again from your description it sounds like you had a contact or pressure problem -feel free to review data on our website http://innovationcooling.com/ I can only troubleshoot a problem so much from a distance as you have only anecdotal info - Was your mount squarely mounted? how many psi? is your sink and IHS flat? etc. 99.99% of the users can not supply me with that info and I can only point to likely solutions. Yes if you want better performance you may have to lapp your IHS, and yes you may have to mod your heat sink bracket if you want better performance.... Blame Intel or the sink manufacturer not the paste

If you want better performance improve your contact and pressure - any paste, any sink....

[RSDXzec]

cool, ive never seen an administrator post.

Hey is there any chance of an offer like this again because i only joined almost 2 months ago(i think)

[dicko555]

Originally Posted by trents

I tried ICD7 recently as it came free with the TRUE I won on an ebay auction. It seemed to give about the same temps as AS5. What I didn't like about ICD7 was its extremely high viscosity. It was like trying to spread road tar or bondo that was almost set up. If you use it you have no choice but to apply a pea-sized dollop on the cpu as the manufacturer web site instructs you to do and then clamp the cooler base to the cpu. You really can't smear it around to make sure you get complete coverage of the mating surfaces like you can with AS5. When I removed the cooler a few days after trying the ICD7 I could see the compression only gave me about 75-80% coverage and it was uneven. I don't know, maybe the free ICD7 that came with the TRUE was used and dried up or something. I went back to AS5.

With regards the difficulty you had with the application.
When applying the ICD 7/24 sometimes it worthwhile warming it in some hot water prior to application this helps with the application. Just put the body of the syringe in the water for a few minutes, not the plunger. Then apply to the CPU as per application instructions and attach the cooler ASAP, so the ICD7/24 is still a little more viscous /fluid. This makes it easier to apply as it is very dense because it is bulk loaded with 92% micronized diamond particles and this just makes it a little easier to work with. I do this and it makes it very easy to use and you end up with a nice even coverage..

[JoeC]

Originally Posted by trents

You really can't smear it around to make sure you get complete coverage of the mating surfaces like you can with AS5. When I removed the cooler a few days after trying the ICD7 I could see the compression only gave me about 75-80% coverage and it was uneven. I don't know, maybe the free ICD7 that came with the TRUE was used and dried up or something. I went back to AS5.

You should also read our Applications notes your description could be figure 7. As long as the mating surfaces are // and flat as demonstrated in the glass slides there is no simpler fail safe application method than the compression method. Spreading as you describe can introduce air and initiate early compound failure as shown in the slides Not only with IC diamond but with other compounds as well - just because you are used to smearing the paste around does not necessarily mean it is the best application method.

[JoeC]

Originally Posted by dicko555

With regards the difficulty you had with the application.
When applying the ICD 7/24 sometimes it worthwhile warming it in some hot water prior to application this helps with the application. Just put the body of the syringe in the water for a few minutes, not the plunger. Then apply to the CPU as per application instructions and attach the cooler ASAP, so the ICD7/24 is still a little more viscous /fluid. This makes it easier to apply as it is very dense because it is bulk loaded with 92% micronized diamond particles and this just makes it a little easier to work with. I do this and it makes it very easy to use and you end up with a nice even coverage..

Cup of hot water works OK. But I find it unnecessary, I get good results by placing the tip at a 45 Degree angle on center contacting the IHS and pressing the plunger, the paste forms a bead so you avoid coiling or roughing up the applied grease. Each to his own

[trents]

Thanks for the additional input, Joe. That first table with the thermal resistance vs. pressure is pretty revealing. Makes me think the heat sink manufacturers would do well to give more attention to their mounting mechanisms. One of the problems I run into is keeping the heat sink level as you lower it into the bead of thermal compound. If you tilt it one way or the other it will skew the spread pattern. I tried to get my wife to help me eyeball it last time but it's hard to get two heads down into an open computer case - at least with a mid tower anyway.

[JoeC]

Originally Posted by trents

Thanks for the additional input, Joe. That first table with the thermal resistance vs. pressure is pretty revealing. Makes me think the heat sink manufacturers would do well to give more attention to their mounting mechanisms. One of the problems I run into is keeping the heat sink level as you lower it into the bead of thermal compound. If you tilt it one way or the other it will skew the spread pattern. I tried to get my wife to help me eyeball it last time but it's hard to get two heads down into an open computer case - at least with a mid tower anyway.

I agree, some manufacturers I believe hurt themselves on the top performance category by being a little light on the pressure side or not paying enough attention to optimal heat sink base flatness/finish

[JoeC]

Happy new Year all,
Final results. OCAU of the 44 who tested Approximately 80% saw a positive result with an overall average improvement of 3.28C. The 20% who did not from our previous Contact and Pressure testing 80% of the 20% were resolved to be a contact or pressure issue that inhibited optimal results.

All in all a pretty good test run and line up well with the total group averages from the other forums.



[IMG][/IMG]

[Blazing fire]

Has any of you tested mx-2, 0098 and this thermal compound? I would like at least 5 reseats to make it a fair test.

[JoeC]

Originally Posted by Blazing fire

Has any of you tested mx-2, 0098 and this thermal compound? I would like at least 5 reseats to make it a fair test.

Check out the previous page there are some website review comparisons or you can go to innovationcooling.com Homepage for a couple comparisons. Problem is it is not widely used enough to get a great comparison I would like more, as is I have 5 @ OCAU with about 2 C average advatange for ICD



User- Comparison paste-Result yellow MX2
EC MEISTER *MX-2 -2
buzzen * MX-2 -2.25
Lardman
CPU -1
GPU1-6.1
GPU1+.5

[JoeC]

We are running another giveaway @

http://www.overclock.net/freebies/44...-compound.html

If you have an account there you qualify for those that missed the last one -going real fast

Ill' update results here when the results come in.

[JoeC]

OCN or Overclockers.net is starting to post results for those that have an interest

[Hazaro]

Don't forget to tell them you can polish stuff with it

[JoeC]

Latest Beta test results coming in @ Overclockers.net for those that like paste comparisons

[JoeC]

Originally Posted by Hazaro

Don't forget to tell them you can polish stuff with it

I can not find the pics or I would post them

[Mr.Guvernment]

for the life of me i can not find my IC7 or AS5 when i need it, since i changed my heatsink! argh, otherwise i could send in new results.

[JoeC]

[JoeC]

From Overclock.net

User- Comparison paste-Result AS5
jacksknight -1.3 C
Enyalius -1
hiiyah777 -4.5
tofunater "0" change
trexxcrap -7 **idle only reported?
chink -2
Sukach -4C
BIGWORM -6
KarmaKiller -3c
xXkeyboardkowboyXx -11
rx7racer -.91 C
Danbeme32 -7C
intelfan -0.5C
theCanadian -3
XxBeNigNxX +3 C
P£P§! -5
dicko505 -3.75
pun3D -9
spaceballsrules -7
bremaine -2
Microsis -4.5
BlackOmega -7
catmmm -4C

User- Comparison paste-Result mx-2
Jeffmizrahi -4c
ljason8eg +2.5
bulmung -5c
dicko505 -2.75
DIRTYDUCK -1

User- Comparison paste-Result Misc. Compounds
Enjoi -5C * Geil
Yomny -2 * Ultra Gel**
Emperor -8.5 Zalman grease
dicko505 OCZ -1
dicko505 ceramique -2
shnoops + 4 Shin Etsu x23
error10 -7Shin Etsu G-751
TnB= Gir +1 OCZ Freeze
nitd_kim -5 stock compound
tindolos -1 TG-1 No ambient reported
spaceballsrules -3.5 Noctau NT-H1
spaceballsrules 0.0 C - OCZ Freeze
FlAwLeSS_666 -2 Shin Etsu G-751
HaXXoR -7 Stock HP Paste
Polska -11 C ZM - STG1

[dicko555]

Originally Posted by JoeC

I can not find the pics or I would post them

Here is a link to someone who polished their CPU with IC Diamond.
http://www.overclock.net/5454942-post208.html

And here are the latest results from Overclock.net

[erasmus372]

Here are some belated results for you sir, I have ICD7 applied in a 5.5mm dot on a stock i7 920, using a Thermalright Ultra90 and TR1366 mount.

Cores 1/2/3/4

Idle = 28/28/23/23

100% Load on 8 = 59/59/53/53

I will be trying some overclocking later and will post those temps right here.