Date:
Tuesday , December 18, 2007
Author:
Marc Adams
Editor:
Kyle Bennett
Google +1

GeForce 8800 GT Mega Heatsink Roundup

Many of you know that NVIDIA's 8800 GT can run rather hot under load. Our GTs ran at over 90C. That spells disaster in a hot case. Let HardOCP show you the way to cool down that white hot GT while not breaking the bank and getting great performance.

continued...

SYSTEM SETUP

All testing of the various heat sinks will occur on a GIGABYTE P965 motherboard paired up with an Intel Core 2 Duo E6300. We will test each heat sink at the GPU's stock settings of 600/900 and its overclocked settings of 700/1000. This will give us a clear picture of how well these heat sinks scale when the heat gets turned up.

Article Image

Test Methods

For this article the CPU will be kept at stock speed to keep any excess heat away from the video card that could impact the results. Thankfully the application we are using, ATiTool, requires little power from the CPU to load up the GPU core to the max. This will further assist us in isolating the video cards temperature from the rest of the system.

Testing each cooler will be done under very strictly controlled conditions to ensure an even playing field. Zalman's STG1 thermal paste was selected again for this roundup for a few key reasons. Firstly, the thermal paste has been shown to require no curing time allowing us to test each heat sink with equal results. Any curing time could have introduced variables into the equation causing at best dubious results and at worst unreliable ones. Secondly, thanks to the unique applicator of the STG1 it is very easy to apply a thin layer across the GPU consistently thereby removing yet another possible variable from the mix.

Ambient temperature will be kept at 25C for the duration of the tests and measured with a MicroTemp EXP non-contact infrared thermometer. Any variance greater then 0.2C will halt the testing until temperatures return within spec for fifteen minutes. Temperatures for the GPU will be measured using RivaTuner 2.06 and cross-referenced by a Sperry Digital 4 Point thermometer. Additionally, we will test the temperature of the PCB directly behind the GPU as well using the MicroTemp EXP non-contact infrared thermometer.

For each heat sink tested, we will measure the GPU's temperature at the fan's lowest and highest setting should the fan be adjustable.

Idle- Idle temperatures will be recorded after a fifteen minute period of inactivity. Any fluctuation during the last sixty seconds will reset the timer for an additional five minutes.

Load - Load temperatures will be recorded after a fifteen minute period of 100% load. To obtain this we will be using ATITOOL. This has been shown to place a constant and consistent full load on the GPU. The benefit of this is that we will be loading the GPU in a repeatable manner. Any fluctuation during the last sixty seconds will reset the timer for an additional five minutes.

Sound - Sound levels will be measured with a Reliability Direct AR824 sound meter from a distance of one foot away. On top of the video card being outside of a case, being so close will again give us a worst case scenario here. With everything turned off and the room completely silent the meter registered a sound level of 38dB(A). This is a very quiet office where a simple pin drop can be heard.