RadeonSI With OpenGL 4 Showing Nice Performance Against Catalyst

[dungeon]

I played with Bioshock yesterday with Catalyst and radeon... so i can say just - both missrender by default.

Radeonsi needs '--eon_disable_arb_copy_image' to render proper, while Catalyst 15.7 needs '--eon_disable_catalyst_workarounds' and nvidia i guess needs no switches but i am not sure for that one

So nope, those eONs can't be compared across drivers by default properly.

[bakgwailo]

I played with Bioshock yesterday with Catalyst and radeon... so i can say just - both missrender by default.

Radeonsi needs '--eon_disable_arb_copy_image' to render proper, while Catalyst 15.7 needs '--eon_disable_catalyst_workarounds' and nvidia i guess needs no switches but i am not sure for that one

So nope, those eONs can't be compared across drivers by default properly.

I think the point is they should be - and both versions of the AMD driver does work/perform by default. Shows more a lacking of the driver than eON. At least radeonsi is just missing an OpenGL 4.3 extension though (only Intel currently has it implemented).

[dungeon]

I think the point is they should be - and both versions of the AMD driver does work/perform by default. Shows more a lacking of the driver than eON. At least radeonsi is just missing an OpenGL 4.3 extension though (only Intel currently has it implemented).

No, those are just eON wrapper switches. That is just what should be done by benchmarker before he start benchmarking. He needs to make sure everything is at least visually correctly emited on the screen and not under any condition to benchmark mess on the screen.

With radeonsi currently if someone run phoronix Bioshock benchmark and don't apply game switch or not run it under mesa env var... benchmark will still run fine with rendering completely broken and he will see huge numbers maybe which means nothing. Same goes to Catalyst without a switch (another one) there are rendering artifacts and benchmark also run slower.

[jrch2k8]

i believe AMD should drop in catalyst(Next Version) support for Desktop GPUs and focus that driver only on RHEL with FireGL cards, move some catalyst developers out into the OSS drivers and focus on that for Desktop users. I mean Radeonsi already support most features the hardware offers(UVD,VDE, DPM, HDMI/S, DP, 2D accel, etc.) better than catalyst.

Only missing items would be OpenCL 1.2+(should be close and 2.0 should be there when HMM lands) and Opengl 4.3+(4.2 is 99.99% done since the last feature has a patch in the works) that is pretty damn close(probably after compute shader 4.4 is just done too and most of 4.5 if not all)

I believe in the long run this move will pay off the better and it will be cheaper for AMD too(and probably will boost radeonsi dev speed) instead on diving the effort in 2 semi complete drivers

[_ONH_]

Well with the 400series as far as id die unterstand AMD amdgpu driver they need toimplemant hw support early to have catalyst bedingt able to use the hw at all, so1-2 8 week in Front of the hw release (Intel des it 2-3 quarter before release if not even more) to have Linux support. So the Situation shouldn't improve til that hw release, the das thing is the pm Firmware bloob should be out at least at release time, to prevent Fiji situation.

[extraymond]

i believe AMD should drop in catalyst(Next Version) support for Desktop GPUs and focus that driver only on RHEL with FireGL cards, move some catalyst developers out into the OSS drivers and focus on that for Desktop users. I mean Radeonsi already support most features the hardware offers(UVD,VDE, DPM, HDMI/S, DP, 2D accel, etc.) better than catalyst.

Only missing items would be OpenCL 1.2+(should be close and 2.0 should be there when HMM lands) and Opengl 4.3+(4.2 is 99.99% done since the last feature has a patch in the works) that is pretty damn close(probably after compute shader 4.4 is just done too and most of 4.5 if not all)

I believe in the long run this move will pay off the better and it will be cheaper for AMD too(and probably will boost radeonsi dev speed) instead on diving the effort in 2 semi complete drivers

That's what they have been doing for years, you can search for John Bridgman's early post in the forum.
They even co-founded the original open driver. It's really cool to see that coming in the long run.

[haagch]

I played with Bioshock yesterday with Catalyst and radeon... so i can say just - both missrender by default.

Radeonsi needs '--eon_disable_arb_copy_image' to render proper,

What do you mean, misrender? I played with my HD 7970M and I thought it didn't look broken: https://www.youtube.com/watch?v=x8F4G4N3xGk

csgo performance feels like it's getting worse over time instead of better...

https://i.imgur.com/mJ9G7TW.png

https://i.imgur.com/AbMkUbr.png

[dungeon]

What do you mean, misrender?

I mean did you run game with '--eon_disable_arb_copy_image' or with MESA_EXTENSION_OVERRIDE=GL_ARB_copy_image ... you need at least one of those so that game render proper.

[haagch]

With MESA_EXTENSION_OVERRIDE=GL_ARB_copy_image. I thought you were talking about misrendering with it.
Well, when you're not, then there's nothing to see here.

[dungeon]

With MESA_EXTENSION_OVERRIDE=GL_ARB_copy_image. I thought you were talking about misrendering with it.
Well, when you're not, then there's nothing to see here.

Well that is my point... it is similar on catalyst, if user don't apply their switch, then it missrender too but differently and is slower.

About CS:GO and Borderlands2, both runs fine with profiles on Catalyst... otherwise performance of those two are similar like in mesa, maybe even lower on some chips