The Best, Most Efficient Graphics Cards For 1080p Linux Gamers
Earlier this week I posted a graphics card comparison using the open-source drivers and looking at the best value and power efficiency. In today's article is a larger range of AMD Radeon and NVIDIA GeForce graphics cards being tested under a variety of modern Linux OpenGL games/demos while using the proprietary AMD/NVIDIA Linux graphics drivers to see how not only the raw performance compares but also the performance-per-Watt, overall power consumption, and performance-per-dollar metrics.
I've ran some similar tests like this before, but now with having the latest drivers, it's time for another go: NVIDIA 355.11 and the brand new Catalyst 15.9. It's similar to the open-source comparison earlier this week but with more hardware, more interesting tests, and using these OpenGL 4 proprietary drivers generally relied upon right now by Linux gamers.
The same Core i7 5960X test system was used while running Ubuntu 15.04 x86_64 with the Xfce 4.12 desktop and Linux 3.19 kernel. The tests run for this comparison included Counter-Strike: Global Offensive, Metro Last Light Redux, UE4 Elemental, UE4 Matinee, and Unigine Valley. In this article we're just looking at the 1080p results given that 1920 x 1080 is currently the most common resolution for Linux gamers. In another article in the days ahead is the same tests and hardware when run at 2560 x 1440; if there's enough interest, I may do one with updated 4K Linux gaming results too.
On the AMD side the tested graphics cards were the Radeon R9 270X, R9 285, R9 290, R7 370, and R9 Fury. On the NVIDIA side there was the GeForce GTX 950, GTX 960, GTX 970, GTX 980, GTX 980 Ti, and GTX TITAN X. The cards tested were based upon the hardware I had available as well as being of recent, modern graphics cards that are still available for easy purchasing from Amazon.com or other major Internet retailers. With focusing on the performance-per-Watt and cost value in this article, the older graphics cards that are harder to find new were left out.
For the pricing information, I used the current list prices on Amazon.com (US) for all of the graphics cards I had purchased or were retail models while the reference samples were listed at the MSRP pricing. The AC system power consumption was monitored using a WattsUp Pro USB-based power meter that automatically interfaces with the Phoronix Test Suite while carrying out all testing plus generates the performance-per-Watt results, etc.
On the following pages are all of these results. If you appreciate all of the work invested into Linux hardware testing and want to see the entire article on one-page, ad-free, consider subscribing to Phoronix Premium.