I have R370 in my laptop. So really would like to see benchmarks on older hardware.
PS. Was a vsync disabled in catalyst driver in this benchmark? I do not see this information in article. Without vsync disabled it is essentially cheating on mesa side.
Do the Catalyst drivers support "swap buffers wait" ?
If yes then we shouldn't be comparing against Gallium with "No SBW".
apples to apples
@elanthis: As an amateur myself I really look forward to a guide/article about that matter as most of the courses I've taken only take up pipelining or that mindset of threading things. But in most cases they don't teach threading at all!
I've also saved the post for future reading
The benchmarks clearly show all the open source drivers are substantially slower than the proprietary ones, apart from a couple of corner cases.
To say otherwise is dishonest tomfoolery.
When is everyone going to wake up and realise that the radeon devs are jerking everyone off ?
To some, this will come over as trollery. To some it will come over as a stuck record.
Well if the devs actually did what they promised, there could be no complaints. But all we get here is the same tired CRAP time after time: the devs are part-time, the proprietary driver has more resources, blah fscking blah.
A CHILD could do a better job than the buffoons currently driving the development work in the open source ATI arena.
Please do it, then.A CHILD could do a better job than the buffoons currently driving the development work in the open source ATI arena.
You seem so well suited for the task.
We should all be thankful that we can watch videos without tearing and play some games on Linux.
The oss drivers are not that bad, just for lowend chips the performance it not even enough to use google earth smoothly - fglrx was much faster there. A google earth benchmark would be interesting. xv usually works and kde/compiz desktops effects but more...? Desktop users just replace the card for games but what should laptop users do? Use older os with fglrx or buy a new laptop?