Half-Life 2 On Wine Is Faster On AMD R600g Over Catalyst

[FireBurn]

What a useless bunch of crap! Why Michael decided to post this is beyond me. He could have at least took the results and reformatted them to make some sense in the graphs

So far there have been more comments on the terrible graphs than when they're supposed to be displaying

[smitty3268]

ati-glsl and ati-arb are wine settings.

just look at the far right for the catalyst results, all the others are r600g.

[AJenbo]

I very doubt that, either. In clear_d3d test text below is saying r600g has much better performance than Catalyst, but all see is virtually overlapping grapths. WTF?

The two graphs are not r600g vs Catalys, it's two modes in Wine, the Catalyst is the spike at the end of the graph, the rest of it is r600g on different versions of Wine.

[Drago]

I very doubt that, either. In clear_d3d test text below is saying r600g has much better performance than Catalyst, but all see is virtually overlapping grapths. WTF?

Oh, now I see it. Still confusing though :P

[Drago]

I doubt that. r600g doesn't have a shader optimizer, how possibly it can beat Catalyst?

I think that part of the deal is that the 'highly optimized' nature of catalyst is often on a per-app basis. In order to look good in benchmarks AMD has spent considerably money profiling and creating optimized code paths for specific applications. If you are not using a particular gaming engine or application that AMD didn't anticipate or put a lot of effort into optimizing then the open source drivers are going to be more competitive.

[dhewg]

@przemoli, dhewg, Drago
As clearly stated in the article, the Catalyst result is the one at the right end of the graphs. Glsl and arb have nothing to do with the driver. The text makes perfect sense to me...

Oh, now I see it. Still confusing though :P

[drag]

I have noticed this with starcraft2. The open source drivers tend to work better then the catalyst. At least for me.

..

I think that part of the deal is that the 'highly optimized' nature of catalyst is often on a per-app basis. In order to look good in benchmarks AMD has spent considerably money profiling and creating optimized code paths for specific applications. If you are not using a particular gaming engine or application that AMD didn't anticipate or put a lot of effort into optimizing then the open source drivers are going to be more competitive.

[GreatEmerald]

Wait, how are these tests CPU-limited? It's a freaking Core i7! With four cores, 2.8GHz each, and hyperthreading!

[AJenbo]

I'm sorry, but "clearly" is wrong wording. The labeling is very confusing and also: the last two result seemed to be Catalyst, according to the first page, not only the last one.

Yes the two last, as in the blue and the green dot to the fare right. I agree that this isn't the best organised test results, but if you read the artical you will find that Michael wasn't the one who set up this test.

[ChemicalBrother]

As clearly stated in the article, the Catalyst result is the one at the right end of the graphs.

I'm sorry, but "clearly" is wrong wording. The labeling is very confusing and also: the last two result seemed to be Catalyst, according to the first page, not only the last one.