Benchmark Results: Thermal Test
After testing each cooler with its stock fan, we threw on our own 2500 RPM fan to separately evaluate the performance of its heat sink. The second test doesn’t apply to Rosewill’s RCX-ZAIO-92, however, since the since the bracket does not support our reference fan’s 120 mm frame size.
Zalman’s CNPS10X Performa leads the way in as-purchased temperatures, though it appears its win is due entirely to a higher-airflow cooling fan. Corsair, Gelid, and Xigmatek reach identical temperatures with our reference fan attached, in spite of vastly different fin and base designs.
VRM temperature is mostly affected by airflow and fan height, since the lowest fan position allows the motherboard’s voltage regulator to receive the greatest amount of air. Rosewill’s shorter RCX-ZAIO-92 does well here, matching Zalman’s higher-flow CNPS10X Performa.
We ran into a bit of trouble while testing Deepcool's solution: after watching our first motherboard’s voltage regulator go up in smoke, we had to lower our ambient temperature to 10 C to finish testing on a second, identical motherboard. All results reflect measured temperature minus ambient temperature, but the other solutions were measured at a far-more-realistic 20 C.
While initial tests might have caused us to believe Zalman had the fastest fan, Corsair’s A50 edges it out. This chart is shown in reverse order, since higher speeds tend to generate higher noise, a factor we will consider on the next page.
That's why the same system was used for both tests. It's all mentioned in the article, too.
Read the article before posting please.