RadeonSI OpenGL CTS Tests Running ~30% Faster With Mesa 24.1

[Dukenukemx]

That isn't true, it really depends, also Vulkan itself is lower level than even gallium... it should be possible to eliminate gallium from Mesa3d eventually and the driver acutally be faster and better, and lower maintenance. There is something to be said for different code paths but I'm not sure that is really the case since a lot of the underpinnings end up being the same for both paths.

ZINK > Vulkan
and
"Native" > Gallium have the same level of abstraction.

Vulkan is lower, but is Zink lower? No it isn't, and while it's fast, it isn't equally as fast as RadeonSi.

Acutally there are many cases where ZINK is already faster now ... and where it ends up not being faster is where there is a lack of optimisations that lead it to choking at higher FPS on say the 7900XTX, while in the same benchmarks it is faster than "native" on slower GPUs.

In most synthetic tests but when you want to actually use it...

When it does win, it's because the 7900 XTX clearly has a driver issue and is slower than a 7600.

[cb88]

Vulkan is lower, but is Zink lower? No it isn't, and while it's fast, it isn't equally as fast as RadeonSi.

In most synthetic tests but when you want to actually use it...

When it does win, it's because the 7900 XTX clearly has a driver issue and is slower than a 7600.

That's using > signs as arrow you dunce. EG, th level of indirection between ZINK and vulkan and native and gallium are identical.

The main reason ZINK looses is CPU overhead, and that is mainly because its new and had few people work on it. If you stop working on the "native" version and start fixing things in the ONE... ZINK implementation for all hardware, you quickly end up with ZINK being faster everywhere.

It's pretty common for the native opengl dirver to ALSO have slowdowns just like you are seeing with ZINK... in other words you are just cherry picking a result that doesn't demonstrate anything at all towards your argument.