[H] Catalyst progress report

[Orangey]

Here is something I and others have been asking for:

http://hardocp.com/article/2010/07/01/amd_ati_catalyst_driver_performance_comparison

Compares the launch driver for 5800, 9.10 and 10.6. Min frames seem to have improved most from a quick glance.

 

[Freddie1980]

](in relation to Crysis Warhead) Well, we just aren’t seeing much improvement here. It seems this game is pretty much tapped out for performance potential

So true, AMD according to he's own numbers have only managed to increase the minimum framerate by a mere 40%, lol. He may know a thing or two about video cards but the guy can't read numbers to save his life.

It’s an interesting comparison though and just shows what can be done with software.

 

[Monkeynut]

So true, AMD according to he's own numbers have only managed to increase the minimum framerate by a mere 40%, lol. He may know a thing or two about video cards but the guy can't read numbers to save his life.

It’s an interesting comparison though and just shows what can be done with software.

The average has gone up by a mere 1fps though.

 

[mmj_uk]

I like how he picked the few games that they for optimize every month, what about the thousands of other titles people play? a bit of variety wouldn't go amiss.

 

[RavenXXX2]

Check out the rez they are using, performance comparisons should be spread over a number of resolutions, the one they use is to high for a 1GB card TBH. Dirt 2 shows way bigger numbers at lower resolutions. Just goes to show, all these Nvidia supporters banging on how ATI have had months to optimize their drivers and that's why they are so close to Nvidia, when in fact there is hardly any performance increase from release drivers barring Dirt 2, the cards perform superbly as it is, there will only ever be specific game improvements.

 

[straxusii]

The average frame rate is worth next to nothing compared to minimum frame rates.

 

[Helios1234]

http://www.vortez.co.uk/contenttell...catalyst_104_driver_performance_review,1.html

 

[NathWraith]

Here is something I and others have been asking for:

http://hardocp.com/article/2010/07/01/amd_ati_catalyst_driver_performance_comparison

Compares the launch driver for 5800, 9.10 and 10.6. Min frames seem to have improved most from a quick glance.

Interesting article. Proves that not every driver update is neccesarily a good one.

 

[Schnippzle]

I'm wondering though; depending on how older drivers interact with each other, surely there would be a performance difference when running the 9.6 drivers with (for example) motherboard drivers from now vs back then? Performance difference might only be negligible, but I think this may be something that should have been considered.

Only changing the graphics drivers does remove possible variables affecting the results though, and does somewhat isolate the testing.

Ofc, I could be horribly mistaken

 

[cobhc]

http://www.vortez.co.uk/contenttell...catalyst_104_driver_performance_review,1.html

Page 8 shows 3dmark06 and Unigine, not Vantage.

 

[Helios1234]

Page 8 shows 3dmark06 and Unigine, not Vantage.

Thanks for pointing it out, it's been fixed .

 

[cobhc]

Cool, quite a jump in CPU score between the two drivers in Vantage, especially on the Extreme setting.

 

[Alexrose1uk]

I'd have to echo what a few others have said, average framerate has generally not increased massively, but there's been some good improvements in MINIMUM framerates in several games, which as mentioned is just as important as average, and definately more important than max, because between average and minimum, you can work out just what the overall experience is like.
As an example minimum fps 19--->26, or in Splinter Cell, 16-->25fps is certainly a decent improvement, about 50% increase in terms of minimum framerate.
The review also doesn't really address bugs in the drivers; stability/lack of bugs or glitches are just as important as the actual performance.