Introduction

It has been more than 2 years since our last VGA cooler roundup and much has changed since then; the heat output of current video cards has surpassed that of older CPUs, which makes it now quite a challenge to cool your high end video card silently. While two years ago you could get away with a passive cooled Radeon 9700 Pro, if you attempt to do this with ATI’s latest X1900 series you might end up with a smoking pile of rubble. That bring us to today, I have 6 different VGA coolers on the test bench from Arctic Cooling, Zalman and newcomer Sytrin. Unfortunately Thermalright’s V-1 Ultra arrived after I’ve finished all tests so expect a follow-up article with which pit their offering to the contenders we have here today: Arctic Cooling NV Silencer 5 (Rev. 3) Weight: 428gr Material: AlCu Fan Speed: 2000rpm -- Arctic Cooling Accelero X1 Weight: 262gr Material: AlCu Fan Speed: 500-2000rpm -- Zalman VF900-Cu Weight: 185gr Material: Cu Fan Speed: 1350-2400rpm -- Zalman VF700-Cu Weight: 270gr Material: Cu Fan Speed: 1350-2650rpm -- Zalman Fatal1ty FS-V7 Weight: 270gr Material: Cu Fan Speed: 2050-3100rpm -- Sytrin KuFormula VF1 Plus Weight: 242gr Material: AlCu Fan Speed: 1550-3150rpm -- As you can see in the photos on the left, all coolers where tested inside a case, which brings us to the test setup:

Test Setup

JMke's Test Setup
CPU	Opteron 144 @ 2.7Ghz
Cooling	Arctic Cooling Freezer 64 @ 50%
Mainboard	Asus A8N SLI Premium
Memory	2 * 512Mb PC3200 OCZ
Other	XFX 7800GT GlobalWin 520W Silent PSU Maxtor 200GB IDE HDD

The XFX 7800GT video card was kindly provided by Comtechnology.be



The system was installed inside an Antec Lanboy, a mid-sized tower case made of aluminum, it doesn’t come with case fans, but supports 2x80mm (1*font/1*rear) – I placed a very silent Sharkoon Silent Eagle 1000 in the rear running at 12v.

I placed a dBA (AR824) meter next to the closed side panel at approximately 10cm; without the system powered on the ambient noise in the test room measured 35.6dBA (in the photo below you’ll notice that even the slightest noise of the camera is picked up). When the system is powered on without a VGA fan running the noise increased to 37.8dBA. Some VGA coolers dropped below this noise level making them extremely silent (~2dBA over ambient isn’t particularly bad).



I placed a thermal sensor at the in-take of the case measuring the temperature of the air that’s being pulled into the case, with the case closed the ambient temperature inside the case easily surpassed 25°C with in-take temperature at 23°C. The closed environment leads to higher temperatures for all coolers, testing them inside a case gives real-world results which will differ quite a bit from those tests done outside a case.

Outside a case there is plenty of cool air for the cooler’s fan to blow onto the heatsink, inside a case however the air blowing onto the heatsink will be higher causing higher GPU temperatures (logical) but not all coolers behave the same when changing inside/outside testing scenario. It’s not like you can simply add X°C to all your results and be done with it.

To test the impact I tested the Zalman VF900-Cu and VF700-Cu with the case panel off and measured the difference:



With fans at 12v the VF700-Cu performs quite well with the case panel removed, trailing the VF900-Cu by only ~5°C. When the case panel is installed the difference increases to 7°C. However look at the results from the VF900 with its fan at 5v, with the case panel removed it performs worse than the VF700, but with the case closed it has a small lead!

So choosing to test the VGA coolers inside a case will indeed give a more realistic overview of their performance.

Test Methodology

in-take temperature was measured at 23°C for all tests, but temp fluctuations, different mounting and user error can account up to 1-3°C of inaccuracy in the obtained results. Please keep this in mind when looking at the results. Each heatsink was tested repeatedly; if I got questionable results the test was restarted.

The VGA card was stressed by running Real-Time HDR continuously until temperatures.

Arctic Silver kindly send us their “Lumière” thermal testing compound which has the same color as Ceramique, but only a break in time of 30min!

Arctic Silver’s ArctiClean was used to clean off thermal paste between tests.

The XFX 7800GT was tested at two different GPU Clock speeds, 400Mhz and 460Mhz.

The fan speed was regulated differently depending on the VGA cooler:
- The Arctic Cooling models fan speed can be changed through Rivatuner, I set it to 25% , 50% and 100%.
- The Zalman models use 3-pin fan connectors and I used a 3-speed Acousti fan speed regulator which sets voltage to 12v, ~7v and 5v.
- The Sytrin comes with its own fan speed selector and the unit was tested at High/Medium/Low.

Now that you know what products are in this test and how they will be tested it’s time to go over the installation producers of each cooler, see what’s inside the box and how good they performed ->