AMD Radeon Gallium3D Catches Up To Catalyst For Some Linux Games

[phoronix]

Phoronix: AMD Radeon Gallium3D Catches Up To Catalyst For Some Linux Games

Yesterday I ran the latest RadeonSI Gallium3D vs. Catalyst AMD Linux driver comparison, effectively another round of open-source vs. closed-source GPU driver testing. In yesterday's article it was found the RadeonSI performance is improving a lot but the AMD Catalyst Linux driver remains much faster. In this article are benchmarks of the older "R600" Radeon Gallium3D driver compared to the Catalyst driver with graphics cards from the Radeon HD 6000 series and older where the open-source R600g driver provides support and in a more mature state.

http://www.phoronix.com/vr.php?view=19737

[Veerappan]

Very nice.

There's a few tests that look like r600g is hovering around 50% speed for some reason or another, but I'm impressed by the performance of r600g relative to catalyst in most of these tests.

[Slobodan]

No offence, but who plays those games? Are tests on those games even relevant?

[devius]

No offence, but who plays those games? Are tests on those games even relevant?

Every single time.... Holy crap. I'm going to stop reading comments when it's a GPU benchmark article. Seriously... There should be a banner at the top of each GPU article explaining this shit so that people aren't asking the same thing over and over and over and over and over again.

[matyas]

CPU or GPU

Looking at Phoronix benchmarks, I always keep wondering whether it's the CPU frequency that could in some cases affect the performance divide. Catalyst may be forcing the CPU frequency, while with the OSS driver, the default ondemand governor is used. There was no CPU utilization graph, so who knows.

[wargames]

Every single time.... Holy crap. I'm going to stop reading comments when it's a GPU benchmark article. Seriously... There should be a banner at the top of each GPU article explaining this shit so that people aren't asking the same thing over and over and over and over and over again.

I feel your pain although it's true that no one plays those games.

[Luke_Wolf]

Every single time.... Holy crap. I'm going to stop reading comments when it's a GPU benchmark article. Seriously... There should be a banner at the top of each GPU article explaining this shit so that people aren't asking the same thing over and over and over and over and over again.

What's worse is that it seems that each time he posts one people create accounts specifically to complain about this...

[Ericg]

Looking at Phoronix benchmarks, I always keep wondering whether it's the CPU frequency that could in some cases affect the performance divide. Catalyst may be forcing the CPU frequency, while with the OSS driver, the default ondemand governor is used. There was no CPU utilization graph, so who knows.

OnDemand shouldn't be an issue anymore, I thought Ubuntu finally moved over to the Intel P-State driver?

[Detructor]

I feel your pain although it's true that no one plays those games.

if you two aren't playing Xonotic you are missing out...

[Kivada]

if you two aren't playing Xonotic you are missing out...

Missing nothing. It's a Quake/Unreal Tournament clone, which got old fast when they where new.

You can try and make a case for Tremulous/Unvanquished, but it was always a crap version of Aliens VS Predator 2.

Also, I've never created another account, Larabel is just too lazy to run the tests for games people actually play so that the people that are actually buying the hardware have a good idea of how it performs in real world scenarios.