Skip to main content

Curbing Your GPU's Power Use: Is It Worthwhile?

The Tests, Test Setup, And A Side Note

The Tests

Many of today's tests are the same ones we used in our previous article, AMD's Radeon HD 5000 Series: Measuring Power Efficiency. As an example of a non-gaming workload, we are measuring power consumption during Cinebench's OpenGL viewport test. We’re using the results from Cinebench as a measure of desktop application performance.

We're also conducting tests with CyberLink's PowerDirector 8, making sure GPU acceleration is active for GPU-accelerated video filters. PowerDirector also supports video encoding with GPU acceleration.

Since PowerDirector 8 does not natively support hardware encoding on the GPU using AMD cards, we patched the application with the latest update (3022) and installed AMD’s Avivo transcoding tool.

The last test is H.264-accelerated video playback in CyberLink's PowerDVD 9.

Test Setup And A Side Note

We're using the same hardware setup from our previous article. This way, the results are directly comparable to those obtained previously.

We left the power-saving features of the Phenom II X4 955 BE enabled in the BIOS, and we set Windows’ power policy to Balanced. To lower the base system power consumption even more, we altered Cool'n'Quiet with K10Stat, allowing our processors to run at even lower voltages. Using these settings, our test platform’s base power consumption hovered around 55 watts at idle and 80 watts during H.264 video playback. This is right around the threshold where our platform's PSU efficiency drops off.

Our altered voltages mean this isn’t a stock setup, per se. So, these results are not directly comparable if you’re running default voltage settings. You can see the differences in system power consumption between running with default voltages and our undervolted settings below.

Phenom II X4 955 BE & Integrated Radeon HD 3300
IdleCrysisCS 4Cinebench
Undervolted55 W121 W120 W132 W
Default Voltage73 W160 W143 W195 W

There is one minor consideration to bear in mind as you look at the power measurement results: variation between samples. Variations from one card to another exist. They may be due to the components used, board design, and even the graphics chip itself. Your own testing may consequently look a little different from what we have here.

All power consumption numbers are for a full system, sans monitor. They are recorded with a Watts Up! Pro power meter sampling at one-second intervals.