Cooling modern video cards is becoming more and more difficult, especially with users asking for a quiet cooling solution, which is why engineers are now paying much more attention to the power consumption of new graphics card designs. An optimized fan-profile is also one of the few things board vendors can create to impress with reference designs where they are prohibited from making changes to the thermal solution or components on the card.
For this test, we measure the power consumption of only the graphics card via the PCI-Express power connector(s) and PCI-Express bus slot. A Keithley Integra 2700 digital multimeter with 6.5-digit resolution is used for all measurements. Again, the values here only reflect the card's power consumption as measured at its DC inputs, not that of the whole system.
We use Metro: Last Light as a standard test representing typical 3D gaming usage because it offers the following: very high power draw; high repeatability; is a current game that is supported on all cards; drivers are actively tested and optimized for it; supports all multi-GPU configurations; test runs in a relatively short time and renders a non-static scene with variable complexity.
Our results are based on the following tests:
- Idle: Windows 7 Aero sitting at the desktop (1920x1080) with all windows closed and drivers installed. Card left to warm up in idle mode until power draw was stable.
- Multi-monitor: Two monitors connected to the tested card, both using different display timings. Windows 7 Aero sitting at the desktop (1920x1080+1280x1024) with all windows closed and drivers installed. Card left to warm up in idle mode until power draw was stable. When using two identical monitors with same timings and resolution, power consumption will be lower. Our test represents the usage model of many productivity users, who have one big screen and a small monitor on the side.
- Blu-ray Playback: Power DVD 9 Ultra is used at a resolution of 1920x1080 to play back the Batman: The Dark Knight Blu-ray disc with GPU acceleration turned on. Measurements start around timecode 1:19, which has the highest data rates on the BD with up to 40 Mb/s. Playback keeps running until power draw converges to a stable value.
- Average: Metro: Last Light at 1920x1080, representing a typical gaming power draw. Average of all readings (12 per second) while the benchmark was rendering (no title/loading screen). In order to heat up the card, we run the benchmark once without measuring power consumption.
- Peak: Metro: Last Light at 1920x1080, representing a typical gaming power draw. Highest single reading during the test.
- Maximum: Furmark Stability Test at 1280x1024, 0xAA. This results in a very high no-game power-consumption that can typically be reached only with stress-testing applications. We report the highest single reading after a short startup period. Initial bursts during startup are not included, as they are too short to be relevant.
Power consumption results of other cards on this page are measurements of the respective reference design.
NVIDIA did not provide us with a GTX 970 reference design, so we can only estimate its power consumption (156 W typical gaming, measured). We did so using a custom-design card with NVIDIA's original reference BIOS, since it uses the proper voltages, memory frequency, GPU base and boost clock.
Just like on GTX 750 Ti, NVIDIA's new architecture brings massive power consumption improvements. Non-gaming power consumption has been reduced further, which increases the gap to AMD even more, especially in Blu-ray playback.
In gaming, the GTX 970 shows impressive power consumption numbers that seemed impossible just a few months ago. The card is fast yet incredibly power efficient at the same time. This is, in my opinion, the key improvement in Maxwell, as power and heat are no longer an issue when it comes to scaling up GPU performance. These improvements will also serve as a foundation for improvements in laptop and mobile gaming, where battery life is a crucial factor.
The ASUS board roughly consumes as much power as the NVIDIA GTX 980 reference board, which I find a bit surprising. There are many possible reasons for these results, though, since ASUS uses a different VRM configuration and has overclocked their card considerably. There are also random variations between cards.
Still, the ASUS GTX 970 uses less power than any of AMD's recent high-end cards, while delivering higher performance.