Thermal Compound Comparison

Companies: ArcticSilver, OCZ, GC Electronics, Silmore
Products: Various Thermal Compounds
Street Price: Varies, Generally Around $10
Date Reviewed: March 18, 2001
Reviewed By: Jason Rabel

Test Setup:

For a testing platform I used my ABIT KT7A-RAID which has a thermal probe underneath where the CPU sits. I had to bend the probe up a little to ensure that the thermal probe was making actual contact with the bottom side of the CPU. For a CPU I used my 1GHz Athlon (courtesy PCNut) overclocked to 1.13GHz @ 1.85v. To take the actual temperature I used Motherboard Monitor 5, and I have a radioshack unit for the ambient air. For the heatsink I used the Vantec FCE-62540D.

I also used my VH6-II which also has a thermal probe mounted underneath where the CPU sits. For an Intel CPU I used my P3-550 (FCPGA) overclocked to 800MHz @ 1.80v. I used the same stuff to record the temperatures & used the same heatsink as the Athlon setup.

Ambient temperature was always maintained between 78.5F - 79F or 25.8C - 26.1C.

To make sure my tests were accurate, I ran the tests 3 complete times on each platform. I cleaned off the CPU core & heatsink each time with WD-40, then rubbing alcohol. The reason for using the WD-40 is because that is one of the few things that actually will completely clean off Arctic Silver, and I didn't want any leftover residue to affect my results.

The idle temperature was taken with the computer sitting at the Windows (98SE) desktop. To stress the CPU I used Prim95's torture test and let it run for about 20 minutes to make sure the CPU reached it's maximum temperature.

Test Results:

After running the tests I was rather disappointed, one of the reasons I ran them so many times was because I couldn't believe the results. All the compounds produced identical temperatures! Here's a quick table, then I will discuss it more in-depth in the conclusion.

Conclusions:

Does this mean that the "premium silver" compounds are worthless? Hell no!

Remember, the tests I conducted were with just normal air cooling, there is only one point where the thermal compound needs to be applied, and it is a very small area at that. If you are into the more exotic cooling with peltiers, cold plates, and waterblocks, then perhaps you will see where the silver compounds have the edge considering you have several layers of compound over a much greater surface area. Still in the end, most other reviews I have read have shown little to no improvement using the silver compounds. It all just depends on your circumstances if the silver compounds will help or not.

Also remember the consistency and life expectancy of the silver compounds is much greater than regular silicone. I have used both silicone and Arctic Silver for quite some time now, and I know from experience that the silicone will start to harden to an unusable thickness (I'm talking over a long time period like 6 months to a year), whereas the Arctic Silver remains unchanged.

I know that I will keep using the "premium" stuff, simply because it's a better designed product. While I might not see any benefits immediately, over the long term the silver compound is much more stable and reliable than silicone compound.

Pro's:
  Arctic Silver II & OCZ Quicksilver Are Reasonably Priced (Under $9)
  Arctic Silver II & OCZ Quicksilver Have Longer Life Expectancy

Con's:
  No Noticeable Performance Gain Under Regular Circumstances

Rating:
  For The Average User: 3/10
  For The Extreme Overclocker: 7/10

Back To The Main Page

Comments? Talk about it in the Forums!