The installation procedure is the same as previous iterations. See my notes on the 14.3 version if you’d like to see change observations there, or see the post on installing the 14.2 version for full instructions.
Some people are apparently claiming to have experienced significant performance improvements with the driver, but Phoronix has yet to validate those results. The good author didn’t benchmark my hardware or anything analogous (only discrete GPUs and low-end AM1 APUs were tested) so I’ll post some benchmarks ASAP. My former flagship APU (the A10-6800K Richland 4.1 GHz model) should be an interesting test candidate under these circumstances.
So far, I hate to say it, but my desktop experience seems smoother (it was getting a little jerky) but it’s hard to say at this point. I don’t expect much variation from previous benchmarks, to be honest. I’ll get back with the scores ASAP.
My latest benchmarks, if you recall, were as follows:
Unigine Heaven v4.0 @ 1920×1080: 6.02 fps
Unigine Heaven v4.0 @ 1600×900: 8.14 fps
Unigine Tropics v1.3 @ 1920×1080: 24.01 fps
Unigine Tropics v1.3 @ 1600×900: 28.73 fps
There was a slight regression due to my motherboard BIOS update (it appears, though my environment is fairly uncontrolled), but that’s where I stood. I was running the Catalyst 14.2 Beta driver (and no one, to my knowledge, reported performance improvement with the Catalyst 14.3 update, which was very minor).
Today’s benchmarks, following the installation of the Catalyst 14.4 Release Candidate driver, I achieved the following results:
Unigine Heaven v4.0 @ 1920×1080: 6.14 fps (+0.12 or +1.99%)
Unigine Heaven v4.0 @ 1600×900: 8.74 fps (+0.60 or +7.4%)
Unigine Tropics v1.3 @ 1920×1080: 30.86 fps (+6.85 or +28.5%)
Unigine Tropics v1.3 @ 1600×900: 38.44 (+9.71 or +33.79%)
Well I am stunned. Those are significant performance increases! As I said, I don’t have an extremely controlled environment, but I don’t think recent updates have entailed anywhere near such performance improvements. I’ll see if Mr. Larabel of Phoronix might be able to attempt to replicate my results, or perhaps a reader could do so?
My machine configurations were as follows:
This is a description of my machine on March 17th when I performed benchmarks as described with the Catalyst 14.2 Beta driver:
And for today’s test with the Catalyst 14.4 Release Candidate driver:
As you can see, there are some potentially important updates (KDE 4.12.3 to 4.12.4, kernel 3.13.6-200 to 3.13.10-200) thrown in with the ones we care about (fglrx 13.35.5 to 14.10.2, OpenGL 4.3.12795 to 4.3.12874), so my environment is not sufficiently controlled to produce precise results, but it sure seems like something worthy of further investigation. I’m sure happy! I’ll be interested to play some TF2 and L4D2!
TF2 Play Test Update:
Wow, the performance improvement is eminently noticeable. The game is much smoother to my eye. What a happy surprise!