Testing

Testing was done over the course of one month with 2 different CPU platforms. Each compound was monitored over the course of 4 days, then removed for another candidate. We mentioned the break-in period for the Nanotherm products (this also applies to the two Arctic Silver products), so this is the primary reason why we're testing for this extended period. The test systems are as follows...

#1) AMD Athlon XP 1700+, MSI KT3 Ultra2, Vantec Aeroflow Socket-A, 512MB PC2700, 120GB Maxtor (RAID configured)

#2) Pentium 4 2.4B, Shuttle XPC SS51, Vantec Aeroflow Socket-478, 512MB PC2700, 120GB Western Digital

The Athlon system runs 24/7, and shutdown only for the time to swap compounds. Since each processor is going to be using one compound every 4 days, this will total 96 hours (give or take time for coffee breaks, etc...) of testing. Prime95 will be running fulltime, stopping only when the PC is needed for other tasks.

The Pentium 4 is the gaming and LAN rig, and will probably be going through the most highs and lows, since we shut down the PC whenever it's in use. This should test the compounds performance in what will likely be the most common user scenario. Prime 95 will be run when it's time to take temperature readings.

The graphs will be presented in two parts. The first graphs will be all the compounds performances the first day of testing. The second set of graphs will be at the end of their test cycle. All temperatures are at full load.

Day 1

AMD XP

Pentium 4

Immediately, we can see there is little difference between the compounds, with the exception of Nanotherm's Silver and EXP products. I was quite surprised to see such a large difference, but the margin was about the same for both computing platforms. Arctic Silver 3 was very close behind, whereas the rest were locked up at about the same temperature.

Day 4

AMD XP

Pentium 4

As with our last roundup, the 72 hours needed for some of the compounds to set-in actually improved performance. I didn't see a drop of more than one degree Celcius though, and I think that's about all the improvement we'll see. I'm actually using Nanotherm Ice 2 (it was the last compound tested) a week after we've recorded these results, and load temperatures are about the same.

I will be quite interested to see the final product of the Nanotherm prototype as it was easily the best performing compound. The Nanotherm Silver XTC competes well against Arctic Silver 3, and the Nanotherm Ice II (and Blue II) are fairly unremarkable considering that performance wise, they were tied with most of the pack.

The generics, this time around, didn't do too well, with the exception of the Vantec Stars-350. This paste was really easy to apply, and not too difficult to clean up, and performed well. Considering it comes in a syringe, is resealable, and best of all, free, makes this one a winner.

Final Words

I've always found it odd when manufacturers claim a 5C+ improvement over other thermal compounds. In our last roundup back in February, we saw that wasn't the case. The only way I can possibly see that much of an improvement is if it's compared to the worst possible compound out there. This time around, although it isn't a 5C difference between the best and the worst, there is more of a gap.

The AOS and generic silicon simply would be classified as the "worst" compounds you can use. Nevermind the performance, but both compounds were difficult to apply, taking me much longer to get that "right" amount on than the other compounds.

The Nanotherm Silver XTC and Nanotherm EXP were two products that really surprised me. Although AS3 may be a little better than past Nanotherm products, the difference was never that great. I generally preferred using Nanotherm, or Alumina because they were easier to clean up. Not only were the EXP and Silver XTC easy to clean up, they really dropped the temperatures to a point where it was noticable.

The Nanotherm Ice II and Blue II were decent products, matching the Arctic Alumina in performance. One thing they have going for it is they tend to cost a little less than Alumina.

The arguement of retail compounds vs free ones is a bit tricky. The benefit of retail compounds is that you can do a little research on them before buying. All the Nanotherm compounds contain no silicon, which as mentioned earlier, would mean more even performance over time. I have used generic silicon before (about 3 years ago) on an old K6 CPU, and recently did a little housekeeping on it. When I removed the HSF, I was pretty shocked to see the compound literally evaporated. Now, I don't think most of our readers would neglect a CPU for that long, but there are people who rarely, if ever change their heatsinks.

In terms of performance, other than the XTC and EXP products, performance between the compounds was relatively equal across the board. Even the absolute worse was only one degree higher than the Nanotherm Ice and Blue. The question now is what to buy and do you need it. Well, my answer is you don't need to buy a retail compound unless you've run out of whatever you just got with your cooler. As for which one to buy, for performance Nanotherm Silver XTC is worth your consideration. In terms of ease of use and cleaning up, the XTC, Ice II (and Blue II), and Arctic Alumina would be my choices. Arctic Silver 3 falls between all of them in terms of performance, but it will cost more, and cleanup isn't as easy.

As for the Nanotherm EXP, it was very impressive, and we hope to have a closer look at the final product when it's released. I can't recommend it at this time since it's not finished yet, but if the prototype is any indication, it's going to be something worth picking up.

All Nanotherm Products are...

Agree? Disagree? Discuss it in our forums

Previous Page - Competition

Home