24-Way AMD Radeon vs. NVIDIA GeForce Linux Graphics Card Comparison

Written by Michael Larabel in Graphics Cards on 27 January 2014 at 10:59 AM EST. Page 1 of 5. 15 Comments.

After this weekend carrying out a 25-way open-source Linux graphics driver comparison featuring AMD Radeon, Intel HD Graphics, and NVIDIA GeForce hardware, the tables have now turned to look at nearly the same assortment of hardware but when using the high-performance, proprietary Linux graphics drivers. We've also upped the demanding OpenGL benchmarks used -- including the Source Engine -- as we see how the AMD and NVIDIA binary graphics drivers are doing to start 2014.

The testing principles for this article were nearly the same as this weekend's open-source GPU driver article but instead focusing upon the two major Linux proprietary graphics drivers: NVIDIA and AMD's Catalyst. Nearly the same hardware was also used except for no Intel HD Graphics since they have no proprietary graphics driver (only open-source!) and being able to test a more diverse range of NVIDIA GPUs compared to the very buggy Nouveau driver situation. We were also able to use more demanding Linux OpenGL games and benchmarks due to these drivers being more performant and boasting greater OpenGL support. The range of graphics cards used included:

- HIS AMD Radeon HD 6450 1024MB (625/667MHz)
- ATI Radeon HD 5700 1024MB (850/1200MHz)
- ATI Radeon HD 5800 1024MB (800/1000MHz)
- Sapphire AMD Radeon HD 6570 512MB (650/1000MHz)
- Sapphire AMD Radeon HD 6800 1024MB (900/1050MHz)
- Sapphire AMD Radeon HD 6900 2048MB (800/1250MHz)
- ASUS AMD Radeon HD 7800 1024MB (860/1200MHz)
- XFX AMD Radeon HD 7900 3072MB (900/1375MHz)
- Gigabyte AMD Radeon HD 8800 2048MB
- ECS NVIDIA GeForce 9500 GT 1024MB (550/400MHz)
- MSI NVIDIA GeForce 9800 GT 512MB (660/950MHz)
- NVIDIA GeForce 9800 GTX/9800 GTX+ 512MB (675/1100MHz)
- XFX NVIDIA GeForce GT 220 1024MB (625/400MHz)
- ECS NVIDIA GeForce GTX 460 768MB (675/1800MHz)
- eVGA NVIDIA GeForce GT 520 1024MB (810/500MHz)
- eVGA NVIDIA GeForce GTX 550 Ti 1024MB (951/2178MHz)
- Zotac NVIDIA GeForce GT 610 1024MB (810/533MHz)
- MSI NVIDIA GeForce GTX 650 1024MB (1084/2500MHz)
- NVIDIA GeForce GTX 680 2048MB (705/3004MHz)
- NVIDIA GeForce GTX 760 2048MB (540/3004MHz)
- NVIDIA GeForce GTX 770 2048MB (540/3505MHz)
- NVIDIA GeForce GTX 780 Ti 3072MB (549/3500MHz)
- NVIDIA GeForce GTX TITAN 6144MB (575/3004MHz)

It's quite a diverse range and goes back to the beginning of where the AMD Catalyst driver currently provides support (Radeon HD 5000 series) and nearly the same for the NVIDIA driver (GeForce 8 series) while going all the way through the current top-end graphics cards. On the NVIDIA side the top-end GPU testing included the GeForce GTX TITAN and GeForce GTX 780 Ti while on the Catalyst side was the AMD Radeon R9 270X. The AMD Radeon R9 290 sadly had to be left out because with the latest AMD.com public Catalyst Linux driver there were Unigine benchmarking issues so we had to use the fglrx-updates driver in Ubuntu Linux for benchmarking instead and there was no R9 290 Hawaii support with that Ubuntu-packaged driver. On the NVIDIA side we were able to use the 331.38 Linux driver without any faults. For those wanting R9 290 benchmarks, you can see last week's article entitled NVIDIA Is Still Killing AMD Over Linux OpenGL Performance -- and the story today is much the same.

All benchmarking occurred from an Intel Core i7 4770K Haswell system while running Ubuntu 13.10 x86_64 with the Linux 3.11 kernel. The benchmarking process was handled fully via the Phoronix Test Suite for providing the best automation and reproducibility. If you appreciate all of this exclusive (and timely) Linux hardware testing, reviews, and driver analysis that are constantly being done at Phoronix please subscribe to Phoronix Premium or consider a PayPal tip. You can also follow us on Facebook, Twitter, and Google+.


Related Articles