OpenGL Performance & Performance-Per-Watt For NVIDIA GPUs From The Past 10 Years

Here's a look at the AC system power consumption data over the course of all the Linux OpenGL benchmarks on the 27 different NVIDIA graphics cards:

Now to make that more consumable, the auto-generated meta performance-per-Watt based upon all of the OpenGL benchmark results in this article:

From the worst (GeForce 8500GT) to best (GTX 980 Ti) is a 72x improvement in the performance-per-Watt based upon all of the test data. From the GeForce 9800GTX to GeForce GTX 980 Ti it's a 6.2x improvement. From the GeForce 8600GT to GeForce GTX 960 is a 25x improvement. Even when comparing a GeForce GTX 460 Fermi to GeForce GTX 960 is over a 2x improvement in the past few years.

Well, if you've been curious how NVIDIA graphics cards have evolved -- or the OpenGL performance in general -- over the last roughly 10 years, hopefully you found all of this test data insightful. The raw performance the performance-per-Watt in going from the GeForce 8 series to the GeForce GTX 900 hardware is incredible. The results show how since Fermi the performance really has been ramping up. It will be interesting to see how NVIDIA's Pascal will fit into this equation with those next-generation graphics cards being just a few months away.

If you want to see how your own Linux graphics performance compare to the 27 NVIDIA GPUs used in this article, it's as easy as installing the Phoronix Test Suite on the distribution of your choice and then running phoronix-test-suite benchmark 1601182-GA-NVIDIA20195. Via OpenBenchmarking.org you can also analyze the results further, normalize the data against a particular graphics card, sort them to your heart's content, etc.

If you appreciate all of the Linux hardware tests done exclusively at Phoronix, consider joining Phoronix Premium today to allow for more of these insightful and unique tests to happen in the future. Alternatively you can also consider a tip via PayPal. This week we are also running a special discounted deal on Premium.