Thermal dispersion coating on CPU heat sink

Status
Not open for further replies.

Finosis

Beta member
Messages
4
I would like feedback regarding a coating that I have applied to my CPU heat sink. This coating is designed to increase the convected heat from a metal surface. The coating is advertised to increase heat flux on a metal part by ~30-40% and I wanted to test and see if this was correct.

The test that I did was using a CPU loading program called 'core damage'. I monitored the heat sink temperature using another program called 'speed fan'. I loaded the CPU to 100% for a period of two minutes and took a temperature sample every 10 seconds during this period. I did several of these tests prior to the coating as well as after. The room temperature during all of these tests was 74 deg F. My CPU is the Athlon 6400 duel core and the heat sink/fan is the 'aero cool x-fire 120 mm'.

The coating seems to have decreased the time to reach its thermal steady state to almost half the time and lowered the maximum temperature by ~30 deg F. I think I'm going to do this to a couple more computers to see if I get similar results. To me the results look pretty impressive but I would like to see what others think about it. If there is anything I'm missing as far as creating as accurate results as possible let me know.

Here is a graph I threw together showing the results:
CoreBenchmark.jpg
 
Now remount with your old paste and test again. Then remount it again and perform another test. Then remount twice with your new paste and test.

Your cpu is old. Were the first tests from the original mount?

As a general rule cpu temps are listed in Celsius. Looks like load temps went from 80C to 62C.
 
Now remount with your old paste and test again. Then remount it again and perform another test. Then remount twice with your new paste and test.

Your cpu is old. Were the first tests from the original mount?

As a general rule cpu temps are listed in Celsius. Looks like load temps went from 80C to 62C.

The mount was not old, I remounted it with new paste about three weeks ago but that is one variable I had not considered until after I coated the heat sink. Either way a one month old mount should not affect the temps too much should they? Ya I converted it to Fahrenheit as those are units I am more accustomed to using.
 
If you were doing it for fun then great job especially getting those temps down.

If you were doing it to prove something then start the remounts. Right now you have the statistical equivalent of going out and shooting 5 out of 10 free-throws. Then drinking a bottle of Gatorade and shooting 8 out of 10.

As for if the month old mount could affect the temps that much, if it was a bad mount then yes it could.
 
If you were doing it for fun then great job especially getting those temps down.

If you were doing it to prove something then start the remounts. Right now you have the statistical equivalent of going out and shooting 5 out of 10 free-throws. Then drinking a bottle of Gatorade and shooting 8 out of 10.

As for if the month old mount could affect the temps that much, if it was a bad mount then yes it could.

Cool, I'll probably be doing this to a few more computers just to see what the deal is but I'll go ahead and give each of the other ones a fresh mount and some new grease prior to the first test.

Thanks.
 
Exactly what thermal paste were you using 1st and 2nd.

Also some paste, like Arctic Silver 5, have a cure in period (50-200 hours for Arctic Silver 5). So you may get 1 result with a fresh application and then a totally different result a week or 2 later.

I would also recommend using LinX as the program to heat up the cpu.
http://www.youwatched.com/datajay/linx(0.64).7z

And Speedfan isn't the best either. RealTemp would be a better monitoring program.
techPowerUp! :: Download Real Temp 3.40
 
Exactly what thermal paste were you using 1st and 2nd.

Also some paste, like Arctic Silver 5, have a cure in period (50-200 hours for Arctic Silver 5). So you may get 1 result with a fresh application and then a totally different result a week or 2 later.

I would also recommend using LinX as the program to heat up the cpu.
http://www.youwatched.com/datajay/linx(0.64).7z

And Speedfan isn't the best either. RealTemp would be a better monitoring program.
techPowerUp! :: Download Real Temp 3.40

Actually, I did use Arctic Silver 5 for both mounts. I'll look into LinX and RealTemp too. Thanks for the info.
 
A few years ago Arctic Silver 5 was the best thermal paste available. It was so good that today people still associate it with excellance. But the truth is Arctic Silver hasn't changed while other companies have been busy with Research and Development. The result is, Arctic Silver 5 is now at the bottom end of the list of recommended thermal pastes.

When considering price - performance - ease of application I alway recommend Tuniq TX-2 :D
Newegg.com - Tuniq TX-2 Cooling Thermal Compound - Thermal Compound / Grease
 
Status
Not open for further replies.
Back
Top Bottom