X-bit labs

Roundup: 11 Fans for Two Super-Coolers and One System Case (page 13)

Category: Coolers

by Sergey Lepilov

[ 04/07/2008 | 07:35 PM ]

Testbed and Methods

We tested all fans in three modes:

• Open testbed: Thermalright SI-128 cooler (with only one fan installed);
• Open testbed: Scythe Infinity (Mugen) cooler (one/two fans installed);
• Closed testbed: ASUS ASCOT 6AR2-B system case (two fans installed for air intake and exhaust) with Scythe Ninja Copper CPU cooler and Arctic Cooling Accelero S1 VGA cooler.

We selected these particular processor coolers for a reason. The thing is that Thermalright SI-128 and Scythe Infinity (Mugen) depend a lot on fan performance that is why we considered them a good platform for fan tests. In this case CPU was heated up using OCCT (OverClock Checking Tool) version 2.0.0a in a 24-minute test with maximum CPU utilization, during which the system remained idle in the first and last 4 minutes of the test. By the way, the new OCCT version heats up the quad-core 45nm CPU by ~7ºC better than the previous version 1.1.1. Scythe Infinity was tested with one as well as two fans.

As for the tests inside a closed system case, we used a passive Scythe Ninja Copper on the CPU and a passive Accelero S1 on the graphics card. So, we measured the processor core temperature, GPU temperature and graphics card PCB temperature. The fans were installed in pairs: one on the front case panel to suck the air in, and another one on the rear case panel to oust the warm air out. The fans were fastened with silicon spindles from Noctua fans. To ensure a fair experiment, we removed the side panel fan completely and covered the opening with a corresponding plate. All rear case slots were also covered with brackets except the one for the graphics card. Here we ran the tests in two modes: loading only the CPU (with OCCT) or loading the GPU and graphics card overall with 10 cycles of FireFly Forest test from a synthetic graphics 3DMark 2006 suite in 1280x1024 resolution without activating any additional technologies improving image quality.

We put together the following testbed for our experiments:

• Mainboard: ASUSTek P5K Deluxe/WiFi-AP (Intel P35), LGA 775, BIOS 0809
• Processor: Intel Core 2 Extreme QX9650 (3.0GHz, 1.25V, 2x6MB L2 cache, 4x333MHz FSB, Yorkfield, C0)
• Thermal interface: Arctic Silver 5
• Graphics card: HIS Radeon HD 3870 GDDR4 512MB / 256bit, 775/2250MHz
• VGA cooler: Arctic Cooling Accelero S1 (passive mode)
• Memory:
  • 2 x 1024MB DDR2 Corsair Dominator TWIN2X2048-9136C5D (1142MHz / 5-5-5-18 / 2.1V);
  • 2 x 1024MB DDR2 CSXO-XAC-1200-2GB-KIT DIABLO (1200MHz / 5-5-5-16 / 2.4V).
• Disk subsystem: Samsung HD501LJ (SATA-II, 500GB storage capacity, 7200rpm, 16MB cache, NCQ)
• Power supply: Enermax Galaxy EGA1000EWL 1000W (a default 135mm ~850RPM fan for intake; 80mm ~1550RPM Noctua fan for air exhaust)

The quad-core processor overclocking with all coolers was limited by the efficiency of the weakest fan (or pair of fans). The results are the following:

• Thermalright SI-128 – 3900MHz / 1.525V;
• Scythe Infinty – 3950MHz / 1.550V;
• Scythe Ninja Copper – 3650MHz / 1.350V.

All tests were performed in Windows XP Professional Edition SP2 operating system. SpeedFan 4.33 Beta 44 was used to monitor the temperature of the CPU, reading it directly from the CPU core sensor. The automatic fan speed management feature (Q-fan) was disabled for the time of the tests in the mainboard BIOS. The CPU thermal throttling of our Core 2 Quad processor was controlled with the RightMark CPU Clock Utility version 2.35.

The ambient temperature was checked with an electronic thermometer that allows monitoring the temperature changes over the past 6 hours. During our test session room temperatures varied between 24.5 ~ 25°C. It is used as a staring point on the temperature diagrams. Note that the fan rotation speeds as shown in the diagrams are the average readings reported by SpeedFan, and not the official claimed fan specifications.

Now let’s get to the actual benchmark results.