X-bit labs

Passive Cooling Solutions for Graphics Cards from Arctic Cooling and Cooler Master (page 9)

Category: Coolers

by Sergey Lepilov

[ 05/31/2007 | 05:29 PM ]

Testbed and Methods

The tests of all cooling solutions were performed in a closed system case only in identical testing conditions. The testbed was built with the following hardware:

• Mainboard: ASUS P5B Deluxe/WiFi-AP (Intel P965, LGA775, BIOS 1101);
• Chipset cooler: Thermaltake Extreme Spirit II (4500rpm, 19dBA);
• Processor: Intel Core 2 Duo E6400 (2133MHz, 4x266MHz FSB, 2048KB L2 cache, 1.325V, SL9S9 Malay, Conroe B2);
• CPU cooler: Enzotech Ultra-X (~1550rpm);
• Thermal interface (for all cooling systems): Arctic Silver 5;
• System memory: 2x1024MB DDR2 Corsair Dominator TWIN2X2048-9136C5D (SPD: 1142MHz, 5-5-5-18, 2.1V);
• Graphics cards:
  • Sysconn GeForce 7600 GT 256MB (128 bit, @670/1584MHz with a -90MHz delta);
  • Sapphire Radeon X1950 GT 256MB (256 bit, 655/1485MHz);
• HDD: Hitachi HDT725032VLA360 (SATA-II, 320GB capacity, 7200rpm, 16MB cache, NCQ)
• System case: ASUS ASCOT 6AR2-B Black&Silver ATX case (with two 120mm fans from Cooler Master installed for air in and out take working at ~1200rpm, ~21dBA; and one 120-mm Sharkoon Luminous Blue LED fan on the side panel running at ~1000rpm);
• Power supply: MGE Magnum 500 (500W) + 80mm GlacialTech SilentBlade fan (~1700rpm, 19dBA).

The tests are performed in Windows XP Professional Edition SP2 operating system. We used DirectX 9.0c (released April 2007), ForceWare 93.71 and Catalyst 7.2 graphics card drivers. The graphics cards were warmed up by running the Firefly Forest test from the synthetic 3DMark 2006 benchmarking suite with disabled anisotropic filtering and anti-aliasing.

The graphics cards temperatures were monitored with RivaTuner version 2.1 utility. Each cooler was tested at least two times with a stabilization period of 13-15 minutes between test cycles. If the results obtained in a repeated test were more than 2^o C different from the previous ones, the tests were performed again until we got fine precision results. The room temperature during this tests session was relatively low: 21^o C.

The average fan rotation speed of all testing participants in different modes is listed according to the monitoring data, and not according to their default specs. Default graphics card cooling solutions are marked as “Stock cooler” on the diagrams below. I would like to remind you that Sysconn GeForce 7600 GT graphics card comes equipped with a non-default cooling solution that looks very much like an all-aluminum Zalman VF700. As for sapphire Radeon X1950 GT, it comes with a pretty typical cooler for graphics cards of this type: a single-slot solution with a fan and a copper ribbed heatsink:

In Auto mode under 3D workload the fan runs at 80% of its potential and is very loud. At 100% the noise is simply unbearable. The system can be considered quiet at 40% of the fan speed that is why besides the thermal measurements at maximum fan rotation speed, we have also included the results at 40% of fan rotation speed.

Besides the default graphics card cooling solutions we have also included the results of Zalman VF900-Cu LED cooler on copper heatpipes. It operates in quiet mode at 1380rpm. Arctic Cooling Accelero S1 and S2 coolers were tested in passive mode and with one 120-mm fan from Scythe (20.9dBA, 49.6CFM). Cooler Master CoolViva Z1 was also tested in passive mode and with one 80-mm Cooler Master fan (2000rpm) installed. It is not a noiseless mode, of course, but the fan is very quiet overall.