Testing The Coolers

Testbed

Methodology

The HIS X1800GTO was flashed to 16 pipelines and overclocked to 612MHz core and 612MHz memory (1224MHz DDR) for the test. The voltages used were 1.200 V GPU, 2.004 mV DDC and 2.004 mV DDQ.

Arctic Ceramique was used as the standard thermal paste during the tests, except for the stock cooler which used its pre-applied TIM from the factory.

As this is a performance-oriented comparison, the stock cooler fan speed was locked at 70% (5100 rpm) for all tests. The Zalman VF900 was tested with its fan at 100% (2500rpm) while the Sytrin VF1 was tested with a 92x25mm Delta fan (model AFB0912HH) at two settings - 3000 rpm (high) and 2000 rpm (low).

The Thermalright V1 was tested with its bundled fan attached to the bottom portion of the heatsink, due to space constraints outlined earlier in this article. However, it was also tested with an additional 80mm NMB fan directing air at the upper portion of the heatsink. This configuration is referred to as Thermalright V1 (Dual) in the charts.

Temperature readings were taken from the Radeon X1800's onboard sensors, as monitored by RivaTuner's monitoring function. Idle temperatures were taken after the system was left idle for at least 30 minutes. Load temperatures were obtained by running ATI Tool's "3D View" function, which had been observed to heat up the graphics card more than any other 3D program available.

All testing was conducted in a closed case. Temperatures indicated in the graph refer to the delta, or difference between the reading and the ambient temperature. This eliminates any variations introduced by fluctuating ambient temperatures and ensures more consistent results.