Testbed and Methods
We tested the new cooling system from Thermalright inside a closed system case. Our testbed was identical for all coolers throughout the test session and featured the following configuration:
- Mainboard: ASUS P6T Deluxe (Intel X58 Express), LGA 1366, BIOS 1606;
- Processor: Intel Core i7-920, 2.67GHz, 1.25V, 4 x 256KB L2, 8MB L3 (Bloomfield, C0);
- CPU cooler: Cooler Master Hyper 212 Plus (Noiseblocker NB-BlackSilent XL2 1500 RPM fan);
- CPU thermal interface: Arctic Silver 5;
- Memory: DDR3 PC3-12800 3 x 2 GB OCZ Platinum Low-Voltage Triple Channel (Spec: 1600MHz / 7-7-7-24 / 1.65 V);
- Graphics cards:
- Diamond Radeon HD 4890 OC 1024 MB/256 bit, 925/4400 MHz;
- Zotac GeForce GTX 260 AMP2! Edition 896 MB/448 bit, @675/1400/2100 MHz;
- Zotac GeForce GTX 275 896 MB/448 bit, 633/1404/2268 MHz;
- System HDD: Western Digital VelociRaptor (SATA-II, 300 GB storage capacity, 10,000 RPM, 16 MB cache, NCQ) inside Scythe Quiet Drive 3.5” silencer and cooler chassis;
- Backup HDD: Western Digital Caviar Green WD10EADS (SATA-II, 1000 GB, 5400 RPM, 32 MB, NCQ);
- Optical drive: Samsung SH-S183L;
- System case: Antec Twelve Hundred (front panel: two Noiseblocker NB-Multiframe S-Series MF12-S1 fans at 820 RPM and Scythe Gentle Typhoon fan at 840 RPM; back panel: two Scythe SlipStream 120 fans at 840 RPM; top panel: standard 200 mm fan at 400 RPM at the top of the case);
- Control and monitoring panel: Zalman ZM-MFC2;
- Power supply: Zalman ZM1000-HP 1000 W (with a default 140 mm fan).
To increase the total system heat dissipation and make it the testing conditions a little harder for the participating cooling solutions we overclocked our quad-core processor to 3.93 GHz (+47.2%) and increased its Vcore to 1.35 V (+12.5%) in the mainboard BIOS.
The memory voltage was at 1.62 V and its frequency was 1500 MHz (7-7-7-14_1T timings).
The testing programs were installed under Windows 7 RTM x64. We used DirectX libraries (from March 2009) and Catalyst 9.8 and GeForce 190.62 graphics card drivers. We used 10 runs of FireFly Forest test from the synthetic 3DMark 2006 suite in 1920x1200 resolution to warm up the graphics cards. We enabled full-screen antialiasing 4x and anisotropic filtering 16x:
Besides, we additionally used FurMark version 1.7.0 stability test that was launched for about 10 minutes in 1280x1024 resolution:
We used RivaTuner v2.24 (created by A. Nikolaichuk aka Unwinder) to monitor the frequencies and temperatures of our cards.
The tests were run at least twice for each tested cooling system. The temperature stabilization period between the two test cycles was about 10-12 minutes. The ambient temperature was checked next to the system case or open testbed with an electronic thermometer with 0.1°C precision that allows monitoring the temperature changes over the past 6 hours. During our test session room temperature stayed around 23.1-23.5°C.
Let’s take a look at the graphics cards, their cooling systems and specifications:
Why did we test our cooling solution GeForce GTX 260 and GTX 275 with pretty similar characteristics? First, because they have different stock coolers (and different PCB layouts). And second, because GeForce GTX 275 that we tested first has no diodes for monitoring the temperatures of VRM components, which are just as important as the GPU temperatures. GeForce GTX 260, on the contrary, does have these diodes. Graphics cards stock coolers were tested in three modes: with the fan in automatic rotation speed mode, at moderate level of noise that was determined subjectively for each graphics card, and at maximum fan rotation speed.
Besides the stock coolers, we also tested Radeon HD 4890 with Arctic Cooling Accelero TwinTurbo in three rotation speed modes of its two fans:
We used this cooler with the same VRM2 heatsink without a fan, just like the T-Rad2 GTX. I wouldn’t recommend leaving voltage regulator components of the Radeon HD 4890 card without any cooling at all.
We have also included a well familiar Arctic Cooling Accelero XTREME GTX 280 into the tests on GeForce GTX.
We tested its cooling efficiency in three fan rotation speed modes too.