Right, by "better heatsink" I meant a properly lapped one. My heatsink was not flat; it had many grooves in it. I guess that explains the test results: the IC compound couldn't do much because the real bottleneck is the poor finish on my heatsink.Relative flatness/finish of the sink base/IHS or mounting pressure are key to extracting the performance out of IC Diamond or any other Compound.
Innovation Cooling IC Diamond 7 TIM Giveaway/Testing
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
Figure #1 is an idealized contact.may012 wrote:Right, by "better heatsink" I meant a properly lapped one. My heatsink was not flat; it had many grooves in it. I guess that explains the test results: the IC compound couldn't do much because the real bottleneck is the poor finish on my heatsink.Relative flatness/finish of the sink base/IHS or mounting pressure are key to extracting the performance out of IC Diamond or any other Compound.
Sink and IHS are perfectly flat. With proper pressure and heat the paste flows and the gap separating sink and IHS closes or resolves down to the average particle size - the size of the particle determines the ultimate bond line thickness (BLT) Smaller the gap the better the performance in conjunction with material thermal conductivity determines performance. My experience is that too much paste will not affect performance -excess is just forced out
Figure # 2 is an illustration of the interference of high spots between the mating surfaces.
If the high spot is the same height or higher as the particle thickness, Particles on top of that high spot will add a whole new layer of compound to the interface, 2X the amount of paste in the BLT increasing the resistance and negatively impacting performance - Flatness is the first priority when seeking best performance.
W/mK Bulk thermal conductivity is only half the equation, the other half being the Bond line thickness or BLT, and contact resistance which relates to how well the compound fills the voids/air gaps.
A grease with low a bulk thermal conductivity might not be so bad if the contact resistance is good and has a thinner bond line.
AS5 for instance. IBM and indigo-xtreme report a range of 0.9 and 1.12 W/mK for AS5. AS5 gains some performance back with a low contact resistance.
Shin Etsu @ 3.6W/mk also has better contact resistance which seems indicated by many independent user performance tests
http://indigo-xtreme.com/comparison.html
http://www.electronics-cooling.com/html/2009_feb_a2.php
=============================
AGENA @ 1150 MHz
Scythe Orochi, Passive
Ambient: 24 C
Arctic Silver 5 (thin spread method)
Idle: 46 C
Load: 53 C
Seven Carat Diamond (bead method)
Idle: 47.5 C
Load: 54.5 C
=============================
KUMA @ 2700 MHz
AMD Stock Heatsink @ 3125 RPM
Ambient: 24 C
AMD Stock Thermal Compound (Pre-applied)
Idle: 26 C
Load: 38 C
Seven Carat Diamond (bead method)
Idle: 27 C
Load: 39 C
=============================
I would like to thank Innovation Cooling and Mike Chin for the opportunity.
AGENA @ 1150 MHz
Scythe Orochi, Passive
Ambient: 24 C
Arctic Silver 5 (thin spread method)
Idle: 46 C
Load: 53 C
Seven Carat Diamond (bead method)
Idle: 47.5 C
Load: 54.5 C
=============================
KUMA @ 2700 MHz
AMD Stock Heatsink @ 3125 RPM
Ambient: 24 C
AMD Stock Thermal Compound (Pre-applied)
Idle: 26 C
Load: 38 C
Seven Carat Diamond (bead method)
Idle: 27 C
Load: 39 C
=============================
I would like to thank Innovation Cooling and Mike Chin for the opportunity.
-
- Posts: 324
- Joined: Wed May 07, 2003 10:35 pm
- Location: Surrey, B,C
I'm not sure what you've done here, but it seems like you've changed the heatsink, fans, and TIM.rpgman1 wrote:i7 920 at 2.8GHz
Zalman 9700LED with Zalam STG-1
Idle: 51c
Load: 78c
Megahalems with 2 S-Flex "G" fans in push/pull
Idle: 40c
Load: 65c
This is a major improvement over my previous temps. Thank you Innovation Cooling and Mike Chin for letting me have a sample of ICD7.
What we need is a direct comparison between any random thermal paste/material and ICD7. Both tests need to use the same heatsink and fan/s, and the fans have to run at the same speed.
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
Keeping it 1 to 1 comparisons is the only way to do it otherwise it nulls the result1398342003 wrote:I'm not sure what you've done here, but it seems like you've changed the heatsink, fans, and TIM.rpgman1 wrote:i7 920 at 2.8GHz
Zalman 9700LED with Zalam STG-1
Idle: 51c
Load: 78c
Megahalems with 2 S-Flex "G" fans in push/pull
Idle: 40c
Load: 65c
This is a major improvement over my previous temps. Thank you Innovation Cooling and Mike Chin for letting me have a sample of ICD7.
What we need is a direct comparison between any random thermal paste/material and ICD7. Both tests need to use the same heatsink and fan/s, and the fans have to run at the same speed.
Last edited by Innovation Cooling on Mon Aug 10, 2009 5:51 pm, edited 1 time in total.
I tried to use my sample on my graphics card over the weekend.
it was so dried up and hard that I could not even squeeze it out of the syringe unless I applied some serious pressure.. i was afraid I was going to break the thing.
Even then, it did not stick to the heat sink, but rather it curled around and stuck to the syringe.
it was so dried up and hard that I could not even squeeze it out of the syringe unless I applied some serious pressure.. i was afraid I was going to break the thing.
Even then, it did not stick to the heat sink, but rather it curled around and stuck to the syringe.
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
Not a common complaint with 100,000 tubes out there, I guess that makes you a special case- I keep a few syringes around from every batch and test them myself, some 3 years with no problem.kittle wrote:I tried to use my sample on my graphics card over the weekend.
it was so dried up and hard that I could not even squeeze it out of the syringe unless I applied some serious pressure.. i was afraid I was going to break the thing.
Even then, it did not stick to the heat sink, but rather it curled around and stuck to the syringe.
Send it back so I can check it and I will send you a replacement
well the tube is empty now -- seems the plastic was strong enough to survive me covering my gpu heat sink.Innovation Cooling wrote:
Not a common complaint with 100,000 tubes out there, I guess that makes you a special case- I keep a few syringes around from every batch and test them myself, some 3 years with no problem.
Send it back so I can check it and I will send you a replacement
trouble is I think the gpu died somewhere in the process, so i wont be able to get any temp measurements.
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
pm your mailing address and I will personally inspect the new replacement sample before it is sent out.kittle wrote:well the tube is empty now -- seems the plastic was strong enough to survive me covering my gpu heat sink.Innovation Cooling wrote:
Not a common complaint with 100,000 tubes out there, I guess that makes you a special case- I keep a few syringes around from every batch and test them myself, some 3 years with no problem.
Send it back so I can check it and I will send you a replacement
trouble is I think the gpu died somewhere in the process, so i wont be able to get any temp measurements.
I would also suggest if you or anybody else has a problem with application send me, or preferably post some pictures in the thread here. A paste pattern spread can reveal a lot about what's going on.
Troubleshooting a problem from a distance with zero info is a little difficult. Don't be shy about providing pics
As noted earlier we have analyzed the no results on multiple forums with independent lab analysis of contact and pressure film tests and nearly all of those were either a contact problem or a mechanical/pressure problem.
And as also noted earlier to make the connection, these two solutions -good contact, lapping sink and IHS along with a secure mount(adequate pressure) are the most recommended thermal troubleshooting solutions posted on any forum.
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
lol, you guys undervolt your systems?thejamppa wrote:Group in here is bit below average probably because we use fans that push less air than average fans other use and with lesser rp's. But still results are impressive. Sure IC Diamond 7 is hard to get out of tube but results speak for themselves. Its fearsome TIM, even with low rpm fans.
-
- Posts: 3142
- Joined: Mon Feb 26, 2007 9:20 am
- Location: Missing in Finnish wilderness, howling to moon with wolf brethren and walking with brother bears
- Contact:
Of course, less heat = less need to cooling. Less power = less heat = less noise. Pretty simple ^^ And those who do not undervolt entire systems just undervolt fans.Innovation Cooling wrote:lol, you guys undervolt your systems?thejamppa wrote:Group in here is bit below average probably because we use fans that push less air than average fans other use and with lesser rp's. But still results are impressive. Sure IC Diamond 7 is hard to get out of tube but results speak for themselves. Its fearsome TIM, even with low rpm fans.
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
thejamppa wrote:Of course, less heat = less need to cooling. Less power = less heat = less noise. Pretty simple ^^ And those who do not undervolt entire systems just undervolt fans.Innovation Cooling wrote:lol, you guys undervolt your systems?thejamppa wrote:Group in here is bit below average probably because we use fans that push less air than average fans other use and with lesser rp's. But still results are impressive. Sure IC Diamond 7 is hard to get out of tube but results speak for themselves. Its fearsome TIM, even with low rpm fans.
Too funny - Less watts = less of a delta temperature rise.
A 100W overclocked compared to an undervolted 50W system?
Poll - How many undervolted their systems for the test?
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
This is the final report for SilentPC IC Diamond Test results.
The group as a whole averaged lower than the overall averages, Which could be a reflection of the communities choice of compounds or that the forum is known to have a tradition of under volting their PC's, in either event those analysis are beyond the scope of this effort which is limited to mostly user raw number comparisons.
Response level was approximately 35% one of the lower to date which could be a seasonality factor due to summer vacations, individual priorities etc.
I appreciate the effort involved in remounting sinks and the time involved in testing and reporting results. The great feed back from the people on the tests has helped us to improve our marketing message and to refine the application and troubleshooting feed back to our customers.
Special thanks to Mike for his organizational efforts in initiating the giveaway and and sample mailing.
Thanks to all that participated and thanks to SilentPC staff for the opportunity.
This thread is not closed, any and all updates, new results are appreciated
Andrew
NOTE ON COMPARISONS
When comparing results more credence should be given to larger sample sizes, smaller less so.
1-5 samples 1 or two tests can flip results either way so usually get thrown onto the miscellaneous group.
6-10 samples may start to indicate trends but can be heavily influenced by outliers and so are lightly weighted other than a general trend indicator
11-20 samples - Starting to develop more of a confidence in the trend direction.
20-30 samples - Confidence level improves.
30+ Samples - High degree of confidence
The laptop tests are from Notebook Review and help illustrate the expanded range of IC Diamonds use in in different applications. The VC/GPU tests across multiple forums I am in the process of breaking out and including those done here will be added as a final chart here at some point. The higher delta temps is due to no IHS and a smaller die contact, providing higher heat fluxes, hence the higher temp differences along with other factors such as compound failure due to the high thermal loads and higher % of generic compounds.
Bell Curve Notes
About half the data is reported in round numbers and approximately 50% of the total numbers were fractional numbers, so to include all numbers in the set I rounded the fractional numbers to the nearest ½ degree. This had a minimum impact on the overall numbers, for instance the mean dropped less than 2/100th’s of a degree.
Notes: IC Sampling Vs. Individual Tests
Innovation Cooling elected to use this more or less unique method to introduce our products as the review cycle runs like molasses for thermal compounds, 10 -12 for ICD in the last 2 years with many comparisons already obsoleted due to new product cycles.
Hardware reviews serve an important function along with observations of user experiences allow individual users to either consciously or unconsciously mentally benchmark results. Our problem was there were not enough reviews to to make comparisons on as compound comparisons are notoriously tedious vs. heat sink or other hardware.
Single tests , individual or even those done by Innovation Cooling are anecdotal in nature subject to limitations of methodology. While most pursue the most rigorous test procedure possible they still encounter fluctuations of several degrees C between tests/reviews.
Why does every test come to a different conclusion? The problem is that their sample size = 1.
Even collecting multiple readings the cluster size is = 1. An individual can collect all the data readings off one system, and will still have almost no statistical power (In statistics this is known as "Intra-Sample Cluster Correlation")because the test set up is dominated by methodology. This is a problem not only in paste reviews, but in other hardware reviews, heat sinks, etc. as well.
In the final analysis methodology ends up defining the final placement of all compounds- All test methodologies fail to take into account things that have a major impact on paste performance. For example the mounting system along with mechanical contact between IHS and sink as evidenced with our independent contact/pressure testing. Variability was very high on the contact results with perhaps 1 in 10 having any thing near what you might call full contact, even on those with lapped components.
In considering pressure related to mounting hardware some pastes perform relatively better under poor mounting, others perform relatively better under good mounting (viscous ones such as ICD). Considering ICD - people that had poor results with ICD had very poor mounting. Once they improved that mounting, ICD did considerably better. Of course, so did their old paste. But ICD improvement >> old paste improvement. Generally, this resulted in ICD>old.
In summary, sample size = 1 tells little. . Sample size matters!
ICD has been extensively tested by 391 independent users in 11 forum groups data that is compiled with real world, real users test results
The group as a whole averaged lower than the overall averages, Which could be a reflection of the communities choice of compounds or that the forum is known to have a tradition of under volting their PC's, in either event those analysis are beyond the scope of this effort which is limited to mostly user raw number comparisons.
Response level was approximately 35% one of the lower to date which could be a seasonality factor due to summer vacations, individual priorities etc.
I appreciate the effort involved in remounting sinks and the time involved in testing and reporting results. The great feed back from the people on the tests has helped us to improve our marketing message and to refine the application and troubleshooting feed back to our customers.
Special thanks to Mike for his organizational efforts in initiating the giveaway and and sample mailing.
Thanks to all that participated and thanks to SilentPC staff for the opportunity.
This thread is not closed, any and all updates, new results are appreciated
Andrew
NOTE ON COMPARISONS
When comparing results more credence should be given to larger sample sizes, smaller less so.
1-5 samples 1 or two tests can flip results either way so usually get thrown onto the miscellaneous group.
6-10 samples may start to indicate trends but can be heavily influenced by outliers and so are lightly weighted other than a general trend indicator
11-20 samples - Starting to develop more of a confidence in the trend direction.
20-30 samples - Confidence level improves.
30+ Samples - High degree of confidence
The laptop tests are from Notebook Review and help illustrate the expanded range of IC Diamonds use in in different applications. The VC/GPU tests across multiple forums I am in the process of breaking out and including those done here will be added as a final chart here at some point. The higher delta temps is due to no IHS and a smaller die contact, providing higher heat fluxes, hence the higher temp differences along with other factors such as compound failure due to the high thermal loads and higher % of generic compounds.
Bell Curve Notes
About half the data is reported in round numbers and approximately 50% of the total numbers were fractional numbers, so to include all numbers in the set I rounded the fractional numbers to the nearest ½ degree. This had a minimum impact on the overall numbers, for instance the mean dropped less than 2/100th’s of a degree.
Notes: IC Sampling Vs. Individual Tests
Innovation Cooling elected to use this more or less unique method to introduce our products as the review cycle runs like molasses for thermal compounds, 10 -12 for ICD in the last 2 years with many comparisons already obsoleted due to new product cycles.
Hardware reviews serve an important function along with observations of user experiences allow individual users to either consciously or unconsciously mentally benchmark results. Our problem was there were not enough reviews to to make comparisons on as compound comparisons are notoriously tedious vs. heat sink or other hardware.
Single tests , individual or even those done by Innovation Cooling are anecdotal in nature subject to limitations of methodology. While most pursue the most rigorous test procedure possible they still encounter fluctuations of several degrees C between tests/reviews.
Why does every test come to a different conclusion? The problem is that their sample size = 1.
Even collecting multiple readings the cluster size is = 1. An individual can collect all the data readings off one system, and will still have almost no statistical power (In statistics this is known as "Intra-Sample Cluster Correlation")because the test set up is dominated by methodology. This is a problem not only in paste reviews, but in other hardware reviews, heat sinks, etc. as well.
In the final analysis methodology ends up defining the final placement of all compounds- All test methodologies fail to take into account things that have a major impact on paste performance. For example the mounting system along with mechanical contact between IHS and sink as evidenced with our independent contact/pressure testing. Variability was very high on the contact results with perhaps 1 in 10 having any thing near what you might call full contact, even on those with lapped components.
In considering pressure related to mounting hardware some pastes perform relatively better under poor mounting, others perform relatively better under good mounting (viscous ones such as ICD). Considering ICD - people that had poor results with ICD had very poor mounting. Once they improved that mounting, ICD did considerably better. Of course, so did their old paste. But ICD improvement >> old paste improvement. Generally, this resulted in ICD>old.
In summary, sample size = 1 tells little. . Sample size matters!
ICD has been extensively tested by 391 independent users in 11 forum groups data that is compiled with real world, real users test results
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
Testing is always relative, you can compare thermal compounds one on one which is commonly done or you can test against absolute limits for a reference point. So I thought some might be interested in another viewpoint on the process.
Notes On The Limits Of Thermal Grease Performance
The best possible performance for any thermal grease would be 100% heat transfer from the CPU to the heatsink – which is impossible. We have measured thermal performance for the best possible case – directly soldering the CPU to the heatsink. In this extreme case using a solder joint, the difference between the CPU and the heat sink was 0.5 degrees C.
Based on test results from 391 users among 11 PC Forums and IC tests, IC Diamond Thermal Compound showed 0.8 – 0.9 C difference between the CPU and heatsink – a difference of only 0.4 C compared to the solder joint.
User results showed other thermal compounds ranging from 1.1 C to 4.7 C difference less performance than IC Diamond, as shown on the performance graph, a difference due to the ingredients in the thermal compound used. Twenty years of thermal compound development have reduced the difference between using a solder joint to about 0.4 C. Further development may reduce this difference by a few tenths of a degree.
In our final market analysis before we launched ICD to market, IC tested the most competitive retail compounds and performed the solder test. These tests gave us the confidence to incorporate the giveaway's for forum testing/market introduction as we were confident ICD would transition well into real world user testing and that at best a competitor may match our performance but will never definitively outperform ICD. IC Diamond and has since been proved out so far with 400+ independent tests on 11 forums with experienced users nationally, internationally in a comprehensive sampling of hardware, software and environmental conditions.
Notes On The Limits Of Thermal Grease Performance
The best possible performance for any thermal grease would be 100% heat transfer from the CPU to the heatsink – which is impossible. We have measured thermal performance for the best possible case – directly soldering the CPU to the heatsink. In this extreme case using a solder joint, the difference between the CPU and the heat sink was 0.5 degrees C.
Based on test results from 391 users among 11 PC Forums and IC tests, IC Diamond Thermal Compound showed 0.8 – 0.9 C difference between the CPU and heatsink – a difference of only 0.4 C compared to the solder joint.
User results showed other thermal compounds ranging from 1.1 C to 4.7 C difference less performance than IC Diamond, as shown on the performance graph, a difference due to the ingredients in the thermal compound used. Twenty years of thermal compound development have reduced the difference between using a solder joint to about 0.4 C. Further development may reduce this difference by a few tenths of a degree.
In our final market analysis before we launched ICD to market, IC tested the most competitive retail compounds and performed the solder test. These tests gave us the confidence to incorporate the giveaway's for forum testing/market introduction as we were confident ICD would transition well into real world user testing and that at best a competitor may match our performance but will never definitively outperform ICD. IC Diamond and has since been proved out so far with 400+ independent tests on 11 forums with experienced users nationally, internationally in a comprehensive sampling of hardware, software and environmental conditions.
-
- Posts: 1839
- Joined: Mon Feb 05, 2007 2:10 pm
- Location: Northern New Jersey
- Contact:
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
Lots of notebook people on the re-paste track these days
http://forum.notebookreview.com/asus-ga ... ess-2.html
http://forum.notebookreview.com/asus-ga ... ess-2.html
From my post at notebookreview
In any event below chart is not final as all results are lumped together, GPU/CPU will be broken out on final.
What is interesting is I would categorize any improvements 10 C and over as catastrophic failure of the stock compound. I count 12 out of 42 roughly 25%
That's a real high number vs the standard PC which are more like 1-2%.
This would translate to early component failure, reduced life expectancy, increased returns/costs of owning a laptop etc.
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
Re: Innovation Cooling IC Diamond 7 TIM Giveaway/Testing
we have been running giveaway's on our new compound Perihelion and the first here from a German giveaway 86 tests and it pretty well held it's own across the board. My first "Blind giveaway" in the sense that my only input was to have my cosponsor post application instructions.
Seems we do better when I keep my mouth shut, go figure
Seems we do better when I keep my mouth shut, go figure
Last edited by Innovation Cooling on Tue Mar 15, 2011 6:31 pm, edited 1 time in total.
Re: Innovation Cooling IC Diamond 7 TIM Giveaway/Testing
no comparisons between ic diamond and ic perihelion?
also: ic diamond user here, very satisfied.
also: ic diamond user here, very satisfied.
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
Re: Innovation Cooling IC Diamond 7 TIM Giveaway/Testing
[quote="fumino"]no comparisons between ic diamond and ic perihelion?
also: ic diamond user here, very satisfied.[/quote]
Posted the final results separated
Kind of have to interpolate between the two tests as we gave out a hundred of each, runs about 1.5 C less than the diamond on avg.
also: ic diamond user here, very satisfied.[/quote]
Posted the final results separated
Kind of have to interpolate between the two tests as we gave out a hundred of each, runs about 1.5 C less than the diamond on avg.
-
- Posts: 40
- Joined: Tue May 12, 2009 9:45 am
- Location: ANSONIA CT. USA
Re: Innovation Cooling IC Diamond 7 TIM Giveaway/Testing
Here is a kind of technical/troubleshooting note
So now at 20 giveaways the most recent from OCUK with something a little different
This set of data that separates OCUK from all the 20 or so groups tested by us.
In the past water cooling is usually less than 5% of the sampled results, on OCUK at this point we are 30% water cooling(marked in red), 70% air.
What made it readily obvious was the statistical cluster of 20% water in the marginal-zero-negative result category. The other 10% is mostly the H50 in the performed as expected group. Might be linked to compound mix but I have a sense that it may have an overall edge in C/P for the H50.
In any event The air cooler group (marked blue) has it over the water with pretty much positive results across the board.
So now at 20 giveaways the most recent from OCUK with something a little different
This set of data that separates OCUK from all the 20 or so groups tested by us.
In the past water cooling is usually less than 5% of the sampled results, on OCUK at this point we are 30% water cooling(marked in red), 70% air.
What made it readily obvious was the statistical cluster of 20% water in the marginal-zero-negative result category. The other 10% is mostly the H50 in the performed as expected group. Might be linked to compound mix but I have a sense that it may have an overall edge in C/P for the H50.
In any event The air cooler group (marked blue) has it over the water with pretty much positive results across the board.
Re: Innovation Cooling IC Diamond 7 TIM Giveaway/Testing
We have been doing a IC Diamond Reliability Survey over at Notebook Review with some interesting results, although only halfway through the survey a major factor in long term reliability relates to Contact and Pressure as we went through the process here with C/P and performance same applies to long term use.
When you have full contact with thermal and mechanical loads in sync with each other fully/evenly distributed across the entire processor and would be a likely canidate for long term use.
When you have full contact with thermal and mechanical loads in sync with each other fully/evenly distributed across the entire processor and would be a likely canidate for long term use.
Just some quick notes
Looking at the data to date it looks to be trending that if you hit a year with no change you are probably clear for extended 1.5- 2- 3 year use. A temp rise above +5C within the first year is a pretty good indicator that you will hit + 20C by year 2.
A kind of a go- no go indicator, those that hit in the 20C range I would call a failure and would repaste and while you have the system apart consider lapping the sink or some mod as TanWare's to improve contact. To note I have a couple of retailers that have agreed to test market the Contact and Pressure indicating film and should retail around $5-$8. Note that we are not selling the product and will not profit from it, we are only promoting it's general use for end users as a tool for improving performance/reliability
I am assuming a 5C +/- error on these tests as most are sloppy about reporting ambients.
As noted in my previous post there does not appear to be any correlation yet as to higher initial temps leading to a failure. Technos started with an initial temp of 58 C and at close to the one year mark ended up 19.25 C while Karamazovmm whose initial temp was 95C ended the year at 90C basically unchanged. There are other examples but this happens to be one of the more extreme. This was a suprise to me as typically notebooks runan average of 15C higher than the overclocked systems so I was expecting some indication in that direction at this point perhaps needs more time.
Applied thermal compound is simple stuff as things go, two mating surfaces with some goop in between and there are not too many avenues to explore to explain a difference in results other than amount of compound applied (we assume everbody gets this one right) Then the contact and pressure between the two joining surfaces.
As we discovered with the C/P testing the weak point in this simple setup is in the contact and is a reasonble cause for good result or bad result or something in between.
Note below yknyong1 contact area being worst case most of the heat transfer would be in the corner/edge area, a significat heat concentration more like a soldering Iron this edge corner/area would be the part of the joint to fail first and as the remaing paste is lightly contacted would then run at higher temps baking out initially then delaminating with further thermal cycling. This kind of contact is a likely canidate for early failure
The following BlazeSempai example at the other end is full contact with thermal and mechanical loads in sync with each other fully/evenly distributed across the entire processor and would be a likely canidate for long term use.
This is as simple as it gets.
yknyong1- Apple Macbook Air13 i5
BlazeSempai p 8700