IC Diamond 24 Giveaway/ Reliability Survey

did a couple repastes first spreading IC Diamond 24 across the CPU and GPU. Initial temps showed a high of 93c @ 81F ambient.

I repasted using the suggestion here

Application

my temps are now a high of 79c high. @ 85F ambient

It didn't seem to turbo as much which is puzzling me. The lower temps are worth it imo though. When not stressing I'm sure games will do ok with the turbo. At least I hope

I used prime95 for 15 min for each stress test

 

Well thats no good, looks like they are way worse. I have seen a few people with similar. I wonder whats going on because I have also seen a few say the temps stayed the same.

When all is said and done after these results I wonder what will happen. Can something be done to make the TIM even better?

 

Guys, where can I go to find a cheap thermometer for my room? Does walmart or the 99cent store have 'em?

 

Thanks for the update - Good stuff at least initially without much data to draw solid conclusions but I can make some guesses.

Notebooks are one of the tougher thermal environments out there, I have no temps losses on PC systems 3 years in the field reported but as I noted earlier notebooks on average run about 10 - 15 C higher than your average overclocked PC.



I had one reported temp increase of 80 to 90C on a GPU that was run 24/7 after a 17 month period.

So from this data and other data I have IC Diamond for laptops it looks like like temps are stable for most in the first 1.5 years give or take depending on ambient temps and use, application, heat sink pressure etc. we have one at the 2 year mark with not much change so some % might extend till 3 years so we end up with a bell curve of results.

Encouraging to me anyway are the anecdotal reports for my competitors having the same issues at 4 - 6 months with temp degradations most likely due to pump out or bake out. This gives us 3 or 4X life expectancy advantage in this application, pretty good in my book

Pump out and Bake out are easily modeled with hundreds of papers on the subject. This is different you guys have come up with something new, well new to me anyway I will probably have to look at epoxy or adhesives to find something similar.

IC Diamond has so little liquid to begin with that it is not really subject to the documented classical problems of pump and bake out.

My first guess is that under high thermal loads for extended periods with ICD basically being a solid thermal cycling degrades the contact with the constant flexing of the joint. This is why I mentioned adhesives as a model as perhaps being more appropriate.

Hopefully we get enough data so we can draw some curves that show something like @ an average of 50C MTBF X years, 60C X years etc. The info so far provides clues for further development and I am going to test one new idea tomorrow

Bryanu ambients even if estimated are helpful and I appreciate you and martee continuing with the installation.

Thanks for updating your results and email me your address and I will send you your tube

 

Here are my temps:

These are the temps from the previous paste that I used... I forgot what it was since I repaste about a year ago and the tube was thrown out.

Ambient for all the tests are at 21 degrees Celsius.

At idle (15 minutes with nothing else turned on):



CPU at LOAD (15 minutes of Prime95 Blend Test):



GPU at LOAD (15 minutes of Furmark Burn-In Test):



Now, for the ICD24 paste results:

At idle (15 minutes with some light browsing):



CPU on LOAD (15 minutes of Prime95 Blend Test):



GPU on LOAD (11 minutes of Furmark Burn-In Test):



Will report back with the same tests in 3 months. If there are anything else that you need, please inform me.

 

Finally got mine today. Will repaste later and run the test.

 

Okay, so I have repasted and I ran Prime 95. Here are my temps

Room temperature = 32'C

Old paste. Min is Idle and Max is Load


Uploaded with ImageShack.us

IC Diamond.


Uploaded with ImageShack.us

Yep, that's all. Will check-in in 3 months.
** Can't reach my GPU so no temps for that

 

I think you guys should look into heat curing compounds. That way you could have a nice thin paste which spreads well with pressure and is nice and easy to work with but as soon as it reaches 50* (or a temperature along those lines) then it would change into a thick gel. That way you could apply easily but still have the resistance to pump out. An important note would be that it should only turn into a thick gel, not a tough adhesive - thermal glue is a complete nightmare.

Another trick which I've heard about but not tried myself (I plan to soon though), if to thin a compound out with IPA or another similar solvent. That makes it spread to give a thinner layer, thus giving better thermal performance. When the chips heat up, the solvent slowly seeps to the edges and evaporates, leaving only the original components behind in a very thin layer.

We've got to the point with IC that talking about improved thermal efficiency isn't really relevant anymore. Thermal pastes are pretty much as good as they can be in terms of performance, where for all intents and purposes the chip and heatsink may as well be one componant. You can only get about 1-2* better than IC at most and that's with liquid metal interfaces which are beyond a pain if you like to swap suff around, not to mention that they're usually aluminium eating.

If you guys focus on making the next IC update (I assume you're going to release a new paste at some point in the future) have the same performance as current IC while making it easier to work with (possibly using some of the ideas above), then you could really rule the market.

 

Liquid in a porous media - No matter how you slice it high liquid content in a porous media (Hand creme type that people like for spread-ability are the primary cause of compound failure) under heat and pressure they pump and bake out. You cant have both easy spread and reliability in the same package which is why OEM's spec bulk loadings above 90% (less liquid) Does not shrink with liquid loss or pump out because it is closer to a solid.

Liquid compounds as a whole get their initial performance from low contact resistance as they fill the voids and gaps, with heat and pressure that performance advantage dissipates quickly.

Depending on use, temps etc initial failure can occur in as little as a few hundred hours

View topic - IC Diamond Results Thread - Maximum PC Forums

Overclockers UK Forums - View Single Post - OCUK Official IC Diamond/ Perihelion Test Results

Notebooks are one of the worst case scenarios for thermal compounds, a kind of accelerated stress test and in this thread alone we had two re-pastes with retail compounds reported with thermal degradation @ 6 and 4 months.

Quality is often measured by durability so if IC Diamond lasts 3 to 4 X as long initially it will classify itself on a quality basis in the final result. Like a cheap rubber tire vs a quality tire, less downtime, more miles traveled, fewer replacements etc.

Easy hand spreading compound is generally not a positive feature to choose in a compound unless that is your most important criteria, If reliability is the major concern then I would seek out compounds that are more viscous in nature


We publish reliability data on our website so someone practiced in the art can look at it and have an idea what we are talking about. Part of the effort here is to translate that idea into the context of real world so end users can apply the same information.

Just to note nobody supplies any reliability information on their products in the retail market other than Innovation Cooling, our competitors are silent on the subject either to their advantage or disadvantage.

 

I'm not suggesting anything about the spread method. I personally use the dot method and it works just fine. However the fact is that the thinner the layer of TIM, the less resistance it imposes. Provided you have enough compound to fill the gaps and bridge the connection then it's all good.

The beauty of thinning out paste with IPA is that it has a low flash point but a comparitively high boiling point. It will evaporate at about 12* but will only actually boil above 80*. Because of this, pumping out from vapours should never be a problem since provided you don't do any high stress activity on your computer for an hour or so after you apply, the temperature will never be high enough to cause cavitation. Since all the solvents would be gone in an hour or two and then afterwards all you would be left with is pure ICD, the paste should never have a chance to pump out since that usually happens over a period of weeks or months.

This is the link I found a few months ago where someone tries and tests the thinning method. Granted, it's not IC but it's still the same concept of a solid material suspended in a thick liquid binder. He managed to shave 1-2* off his temps by thinning the compound out.

Alternative Thermal Paste Application Method - spodesabode.com archive

As for using a heat sensitive binder, I'm not trying to imply that it would be used on top of the standard binder, more as an alternative. In its non cured form, it would be a thin liquid meaning that less has to be used for an application. Then, after the cure cycle has completed the binder would ideally be even thicker than the current one to prevent any pump out from the thinner layer of TIM. Provided it doesn't set like glue then getting the heatsink off shouldn't be a problem. Advantages would include:

-Thinner resulting layer giving reduced resistance
-Less hassle to get the stuff out of the tube and onto the chip
-More potential applications per tube as less would be used per chip
-Even greater immunity to pump out
-Greater ability to conform to the minute crevaces in the surfaces when applied

 

I can not comment directly on IPA as I have not tried it in that fashion but with solvents in general it is not an accepted practice, not sying it does not work I just have not had experience with it.

Problem is prioritizing R&D efforts to a commercial path as it is interesting playing around in the lab it very hard divine what's really going to work on a commercial basis so progress on our end tends to be incremental. It gets expensive following up an avenue of study and hitting a dead end after an extended period.

 

IC Diamond Review on several laptops - I was not really expecting it as I was more focused on the giveaway and generally do not supply samples for reviews. In any event a credible job and doing it on laptops is a noteworthy change from the usual.

Innovation Cooling – Diamond 24 Review | Tech|Inferno

 

Making a heat sensitive binder while retaining the desired properties for thermal paste might be tricky though. If you do it via solvent to get something more viscous, there are the problems mentioned before, let's not forget that the solvent would have to diffuse out of the compound through a very thin layer on the sides. Not respecting a certain curing time before use could be an issue and even then, gas bubbles might get trapped and become an issue even if the solvent isn't past it's boiling point.

If you go for some kind of chemical reticulation, it could work fine if the content of the materials that reticulate are very low, but if they have to be high enough to influence the heat transfer significantly, it might be tricky to find good reactants that don't gimp the heat transfer coefficient of the paste.

Of course, something that can be cleaned with IPA or a not too aggressive solvent would be required too. I don't see myself using THF or dichloromethane to clean a silicon die. I don't know for sure, but i have a feeling those solvents would eat PCB for breakfast given the chance.

Choosing the right materials for thermal paste seems like quite a hassle to me, but a very interesting engineering problem. If investigating the solvent avenue, i'd see FEM simulations as a good starting point .

EDIT: IPA -> isopropyl alcohol
THF -> tetrahydrofuran (nasty stuff that, but disolves a lot of things too)

 

Hey,

I participated in the ICD7 giveaway two years ago with my gateway FX P7801U. The impressive results I obtained back then can be seen here.

(copied from May 2009 post)
Stock 7801u, room temp 20C.
Temps read with CPUID Hardware Monitor, after 2 hours at idle, or 4 hours load using Orthos.

Before:
Idle
- Tzs01 43C
- Tzs02 43C
- Core #0 31C
- Core #1 32C

After
Idle
- Tzs01 39C
- Tzs02 39C
- Core #0 27C
- Core #1 27C

Before
Load
- Tzs01 72C
- Tzs02 71C
- Core #0 61C
- Core #1 61C

After
Load
- Tzs01 66C
- Tzs02 66C
- Core #0 52C
- Core #1 52C

Delta T Idle
- Tzs01 4C
- Tzs02 4C
- Core #0 4C
- Core #1 5C

Delta T Load
- Tzs01 6C
- Tzs02 5C
- Core #0 9C
- Core #1 9C
----------------

I am now updating my results to provide some perspective on the long term effects of this compound.

Background: My laptop is still unmodified. It is run routinely for many hours on end and had a good life as a gaming laptop for the first year I had it. I have cleaned out the heat sync recently to emulate the "like new" condition it was in when the test was first performed.

Setup:
Stock 7801u, room temp 21C.
Temps read with CPUID Hardware Monitor, after 2 hours at idle, or 4 hours load using Orthos (This is the newer version of orthos than what was used previously).

May 2009 Idle
- Tzs01 39C
- Tzs02 39C
- Core #0 27C
- Core #1 27C

July 2011 Idle
- Tzs01 37C
- Tzs02 37C
- Core #0 23C
- Core #1 24C



May 2009 Load
- Tzs01 66C
- Tzs02 66C
- Core #0 52C
- Core #1 52C

July 2011 Load
- Tzs01 82C
- Tzs02 82C
- Core #0 69C
- Core #1 69C

Thanks for the opportunity to share my results.

 

Here is the IC Diamond repaste results, so im off the Wall of Shame for now. xD
My current reaction is how great the temp reduction really was, just check the FurMark temps after repaste.

IC Diamond idle temps.


IC Diamond prime95 stress test on all 8 cores.


IC Diamond Furmark Xtreme Burn-in on HD 6970m.

 

ON MY HONOR IN EXCHANGE FOR FREE COMPOUND I DO SOLEMNLY SWEAR TO TEST AND POST MY RESULTS and to AND TO UPDATE THOSE RESULTS EVERY 3 MONTHS

 

I finally recieved my notebook this easter, so the thermal paste is not that old at all. D:
What i did notice that the stock paste was maybe overapplied.

I followed the steps as linked in the notebook IC Diamond link.
Applied it as a line on the CPU and a pea on the GPU.
Kept the thermal pads and cleaned my fans.
Not much dust either but lower temps sure is great.

 

I know it is counter intuitive but I tested over application amounts with a couple of glass slides some binder clips (60 PSI) and a mock copper IHS. With a micrometer at timed intervals so in 2 hours it was 95% of the BLT (bond line thickness = the average particle size) and overnight 98%. So this is where we get our approximate cure time excess just squishes out till the compound resolves down to the BLT.

We always recommend a little excess as the cost in compound/time/aggravation/ is not worth the quarter savings when you have to redo the mount.

So here I was taking a look at some published information on reliability linked below
and a rough analysis might be with their 30% failure rate in 3years approx 1/3 of that due to breakage and about 2/3 or 20% of all laptops is a component failure probably due in large part to high thermal loads - Heat is number one enemy of electronics.

DEagleson what interests me is when I see 10-15-20 C performance gains on a re-paste in excess of what I would expect over a stock compound especially in a short time frame of 3 months. I see it often enough so unless there is a chronic sink attachment problem out here I am not aware of stock compounds seem to be the weakest link in the chain here.

After a re-paste on a PC in the overclockers forums you have an average of 61C and the laptops here so far average 71 C after a re-paste. I think the extra 10C is the dividing line, in the PC market you may see a few degrees decline and here and there you get the rare 15 -20 C degree improvement but generally most of those are mount related issues unless the compound has been installed a couple of years or more.

The PC failures were generally less than half laptops. The prevalence of early temperature creep in laptops is quite surprising along with the relatively large number of 85 - 90 C systems reported---- 25 to 30 C higher than the PC's, if you have a weak component those kinds of temps have a good chance of killing it. As a kind of general recommendation to avoid being the 1 in 5 failures and as a good maintenance procedure has been suggested here blowing the dust out regularly, I would add a secondary laptop cooler as well as changing the compound every-time it creeps up 10 C past the initial install temps. The cooler it is and the longer time periods it remains cool, the odds that longer system life increases.


http://www.squaretrade.com/htm/pdf/SquareTrade_laptop_reliability_1109.pdf
http://cache.gawker.com/assets/images/4/2009/11/500x_squaretrade3.jpg
Gartner Says Annual Failure Rates of PCs Are Improving, but Manufacturers Can Do Better

 

Received it. Trying to decide if I want to swap my MBP for a W520.

Repaste will ensue after decision.

 

I do force my fans to 100% as stated in my temp readings so the "system" is the same on each test by using "Fn+1".
With stock fan control my temps are around 5c to 8c degrees hotter.

Im so suprised why im getting the great temps, something must have been done wrong with the stock paste.
The GPU is much cooler with FurMark than the first run, where i had to close it after i saw it reach 100c+.

 

hi sorry for being late with the temps have not had so much free time as i wanted have bin at work in Norway at the place god forgot and the devil dident want
"platsen gud glömde og djävulen förlate"ölen utanfor haugesund
i have now Mounted a shims and pasted the Cpu and the Sli 485m GPU's
i used paste on both sides of the copper shims i did three one hour runs with Prime95 set to In -place large FFT's
with the force fan on 100%(Fn+1)
when i saw that the runs where the same i recorded a full run with HW monitor pro
the Ambient Temp Was 22c
and Cpu and Gpu's was all on stock no Oc applied
Then i ran Furmark three times 30 min they where all the same so i recorded it with Hw Monitor pro

CPU Temps One Hour Run Prime95

Spoiler

Here are the Core temps


By torment78 at 2011-07-22


By torment78 at 2011-07-22


By torment78 at 2011-07-22


By torment78 at 2011-07-22


By torment78 at 2011-07-22


By torment78 at 2011-07-22

980x Core temps Idle
Core0 48c to 50c
Core1 49c to 51c
Core2 50c to 52c
Core8 43c to 47c
Core9 51c to 53c
Core10 49c to 51c

980x Core Top temps 100% Load
Core0 98c
Core1 94c
Core2 93c
Core8 92c
Core8 92c
Core9 98c
Core10 98c

GPU Temps 30min Furmark run

Spoiler

Uploaded with ImageShack.us



Uploaded with ImageShack.us

485m Idle Temps
GPU1 33c to 38c
GPU2 32c to 38c

485m Top Temps 100% Load
GPU1 73c
GPU2 74c

Just to inform the Heat zincs for the GPU's have bin Lapped

If you want more info or have question about the cooling modd's just ask i have some pics

 

kizh - gave me a fast turn around on the C/P pressure - Thanks for the extra effort, I will send out for lab analysis this Monday and can email the results perhaps by Friday.

Sensor Products changed their C/P paper earlier this year and I do not quite have the hang of reading the new raw images.

In any event what came back was interesting



The way you read these things is the denser areas as on the bottom 1/3 has higher pressure + more contact/pressure than the the top 2/3's

The upper right quadrant basically has no contact at all.

The poor contact in these areas could be almost anything - rough contour on the sink, a screw mount could be off a thousandths, irregularity in the MB causes the chip to sit at an angle etc. It does not take much to throw things off.

Thermal compound is a mechanical function of Pressure and Contact and most experienced end users are aware of this. Low C/P results in high temps, simple.

kizh's C/P while not terrible will increase the watt density with less contact area the remaining contact area will experience higher loads. This chip is rated for 35W with about 1.3 sq centimeters of area, cut the contact area 30% - 50% on a chip and you will see considerable thermal stress in the remaining contact area much like the 30W soldering iron melting solder.

This kind of an issue will greatly accelerate deterioration of a thermal compound very much like the example below the high contact area in the center dried out while the periphery remained more or less intact. More heat in a smaller area.

Overclockers UK Forums - View Single Post - OCUK Official IC Diamond/ Perihelion Test Results
Overclockers UK Forums - View Single Post - OCUK Official IC Diamond/ Perihelion Test Results

A good guess is that on the 1-5 early failures you will find some % due to poor contact and the ones where the laptop/compound lasts the longest has the best contact.

Might be a good little side business selling laptops that have a certified/qualified contact/pressure report for longer reliability.

I would like to get more C/P samples if any have an interest.

 

Would there be any way we could help out here. If it's just a matter of photo manipulation then I'm sure we can all help out.

 

glad to help, you can post the results here or I will after the email. I'm following the thread.

I challenge anyone to find a thermal paste company searching for answers rather than telling you what you need.

Your business ethic is great in my opinion.

just for those wondering, I'm using a Dell XPS 15 r2. I tightened in accordance to the numbers they graciously engraved into the board.

The C/P test is really easy, especially if you are in the middle of a re-paste. All you do is put the papers between the dry CPU and heatsink. One paper gets thrown away, one holds the image. I would like to see how others compare.

I don't have a scanner, or I could have posted preliminary results faster

 

i dont know where to get the raw papire todo it other wise i would becouse it would give me a good pri lim what needed to be sanded down
if its not flatt make it flatt if some thing is of compesate

 

I don't think my icd24 would still arrive. Its been almost 7 weeks.

 

just PM him with your shipping address and e-mail, he supplies the papers and I think will even include an envelope with return postage.

This is the company that makes the paper
Heat Sink - Surface Pressure Distribution And Magnitude

not sure it would be worth doing on your own, or if they even deal with the public.


Never thought about sanding down the areas that aren't contacting. I rather not sand the cpu, and it looks like I would have to sand 2/3 my heatsink to make the last 1/3 connect. Too much work! Maybe I could find a way to tighten harder on one screw, maybe with a small gasket or something, but I forget which side was up on my C/P test.

 

kizh- your paper went out today for lab analysis.

Anybody that has an interest in doing it it is a great analysis tool especially when linked with temperatures.

Just email me at [email protected] Title it C/P in the subject line and supply your mailing address - simple test two pieces of C/P Paper Clamp between sink and chip then release and you are done

Paste impression pics can be can be a great indicator also


Some background stuff here

Paste impressions:

Poor mounting pressure - thick compound



good mounting pressure - difference between mounts -6 C - light glaze




Another example - Contact paper sample with paste impression of the same mount Light glaze on center indicates good contact and pressure in that area



Another comparison

High contact / high pressure in the area with the thin glaze -

 

Lab analysis provides all kind of detailed info - % of contact area, force applied 2d, 3d pressure visualizations etc.

For example

This is a Dtek FuZion WB C/P from 2008 - Key points, contact area .96 inches and force 43.52lbf and note pressure/contact bias on the bottom half - C/P mostly on the edges of the IHS with light contact and pressure on center - Thermally choked as the heat from center has to cross the IHS to the edges to the sink contact area

 

got my results back, basically its over my head but may be use to others. A LOT of numbers! The actual C/P paper told me all I needed to know. I'd post contents of the e-mail, but it seems confidential between the recipient and sender and I want to respect that. I still encourage people to do the test.

 

Not rocket science as kizh noted. Lots of good info obtained. Raw images reveal much as well as pictures of paste impressions

I cannot find the exact full die size but then have not looked too hard yet but initially pressure is very good as I suspected it would be and contact appears to be around 50% - 60%, 9/10 th's of a centimeter

So basically you have half the area for dissipating 35W - +2X the thermal stress on a compound in an environment that is already warmer than most to begin with, little wonder why people are experiencing accelerated compound degradation.

I currently have 3 people on tap for C/P testing it would be nice to get a broader profile of 20 to 30 and to see how things trend.

 

Plenty of samples left.

I have around 8 - 9 people for the contact and pressure testing I would like to push it up to 20 - 30 any takers?

Just to put this in context as to thermal stress on the compound in the above example 35W on one centimeter in terms of watt density if you were to apply that to a desktop IHS with 8.993cm X 35W = 314.77355 W

If the full 2 cm was in contact you would be looking at approx 1/2 or 17W per centimeter, a more manageable thermal load as far as thermal compounds go.

A little dated but good stuff here - be interesting to see how our tests line up

 

ON MY HONOR IN EXCHANGE FOR FREE COMPOUND I DO SOLEMNLY SWEAR TO TEST AND POST MY RESULTS and to AND TO UPDATE THOSE RESULTS EVERY 3 MONTHS

 

FWIW I have been using ICD7 in my OC'd laptop for over 2 years now, changed out once in that time due to taking the laptop apart. Although the temps are a few degrees higher than day one it seems to be doing okay. Surprisingly it was viscous enough to replace the thermal pads on some of the other components without weeping out the sides where the gap between the all in one H/S and component was quite large (~0.5mm). Easy to use, no complaints. All the best with the newer product.

 

Alright, quick monthly update.

Ambient 26C thanks to colleagues who can't stand "cold" in the office

CPU
Core0: 76C
core1: 77C
Core2: 74C
Core3: 76C

GPU
TS0: 83.0C
TS1: 84.5C
TS2: 83.5C

No change in average delta T for the CPU.
Slightly lower deltas for the GPU.

I'm keeping the Data in an Excel spreadsheet so if you want it at the end of the survey, i'll send the file.

 

On My Honor In Exchange For Free Compound I Do Solemnly Swear To Test And Post My Results And To Update Those Results Every 3 Months

 

While of course I don't have all the graphs etc. the ICD tim is the bomb. I have a severly over stressed P79xx from Gateway. At this point highly modified and running components well beyond the systems designed capability.

Not only with ICD am I able to run these but with 6 months plus of doing so I have experienced no degradation of thermal performance. This is very impressive as the system at its higher overclocked speeds runs very close to component thermal limits.

I have tried other Silver and Ceramic TIM's and inevitably 3-6 months down the line end up redoing the application from thermal performance degradation. This in usually much less demanding systems. Since now exclusively using ICD I ahve never experienced a single issue and can abuse my system with full confidence in thermal performance.

With this I have also used the TIM on my daughters BB Asus U80a. This has been upgraded to a P8400. The installation of the CPU is well over 12 months and again no thermal issues but to be honest it is not an over stressed system.

I should note, while I now have a free tube, all prior installations of the TIM were on my dime! I am just amazed not just by the performance of the TIM from day one but way down the line. The only reason to date for any reapplication is the fact I am an obsesive upgrader. My longest term with one CPU was one year and pulling it appart the TIM on the CPU and open gapped NB had the appearance of having just been done.

In conclusion, and in my experience, ICD is the only high performance TIM that with proper application is set and forget!

Edit, C/P tests can be misleading. You need to apply the TIM on the one side and there are lot of issues. the primary one is letting the system heat up and settle in. With paper there you really can't do that safely. Also while there is never going to be a 100% equal mating what ICD excels at is transfering heat where gaps exist even at no preasure between the chip and sink. This has been well proven by application to NB chips that do not have any contact of the HS to the NB chip.

 

tepes... read the first post...

 

I just re-pasted my new laptop. I also used the contact paper to test compression. I will mail the contact paper out tomorrow and will update with temps in a couple of days.

 

Repasted my GPU and CPU. GPU runs cool (will post results later). Problem is I have Eco Engine modes. And at 0.9 Ghz my laptop runs cooler (~53C) than before repasting (~60C). However, at Eco Off mode (It stays at max: 1.7 Ghz) the temps are just as bad as before (93C if not worse).

It doesn't make any sense. Someone suggested putting the tube in hot water before using it for easy application but I didn't do that since IC website didn't say anything about that. His point was that laptop heatsinks don't put enough pressure. I kinda believe that. But if my repaste was not done correctly how did I get lower temps for low clocks?

Note: I did clean my fan and heatsink in the process.

 

I just repasted my old HP Compaq nw8440 (overheating ftw) because i was cleaning the fans for dust, and this time i did not preheat the paste.
Sure it was a little harder to get the paste out of the syringe but i dont think it should matter that much.

Maybe your temps would be better after some thermal cycles?

 

I reckon it's just turbo boost coupled with a poor cooling system.

 

Are the gaming branded MSI notebooks poor on the cooling side?
I guessed they would at least try to do a good job, sticking a HD 5870m in it.

 

My friend has the GX660 gaming laptop with i7-740QM and a 5870m. It has a button that turns the fans all the way to max, making the fans so noisy they're nicknamed the "twin 747 turbo engines" as there are two fans. Suffice it to say, it never goes above 80 degrees on load with IC Diamond on normal without the turbo fans. Max temps drop to 75 degrees with turbo fans.

 

Hi Andrew, Thanks for the C/P papers which I received today, you have sent me two marked A and B, do I use the second in the IGP or you want me to do the test twice. Thanks.

 

Finally received the paste! Would repaste soon.

 

Sweet.
I hope you get as good temps as i did with IC Diamond repaste.

 

On My Honor In Exchange For Free Compound I Do Solemnly Swear To Test And Post My Results And To Update Those Results Every 3 Months

 

On My Honor In Exchange For Free Compound I Do Solemnly Swear To Test And Post My Results And To Update Those Results Every 3 Months


FurMark 10 Minutes, Max Fans, GTX560m, i7-2630qm:
[Core0 MaxC][Core1 MaxC][Core2 MaxC][Core3 MaxC], [GPU MaxC] @ [Ambient C]

After Install, Sept 2011:
[49][51][46][52], [75] @ [22.2]

3 Months, Dec 2011:
[45][49][45][49], [71] @ [20]

6 Months, Mar 2012:
[][][][], [] @ []

9 Months, June 2012:
[56][58][53][56], [91] @ [26.1]
[47][52][45][46], [74] @ [22.2]

1 Year, Sept 2012:
[50][51][49][51], [72] @ [20]

Dec, 2012:
[][][][], [] @ []

March, 2013:
[47][50][44][47], [74] @ [20]

March, 2015:
[50][51][50][50], [73] @ [19.6C]