NVIDIA vs. AMDGPU-PRO vs. RadeonSI OpenGL Comparison For Holidays 2016

[duby229]

I don't think that explains a potential performance gap.

As we can see from previous performance improvements from the open source driver, games are doing things that are not always optimal, like computing shader outputs that are not used later. By having the driver cleaning the stuff, you get performance improvements. NVidia drivers likely do these tricks from a long time, and probably have more. There's also multithreading that helps when cpu bound. But it's only a matter of time now that the open source drivers get these tricks too.

Sure, I fully agree with you of course. That's certainly a big chunk of it too as can be plainly seen with Bioshock Infinite. When that game launched the drivers were in a state such that it was basically unplayable, but now as can be plainly seen it's performing somewhat competitively. That's what I mean when I say that all AMD really has to do is to achieve that "Good Enough" status on their OpenGL drivers. Once Vulkan proliferates these will become a problem at the game engine level and not the hardware driver level anymore.

[bridgman]

amd dedicated desktop gpu, now are really good, open source driver are amazing.
it's a pity that mobile APU performance are really bad, seems that they are optimizing only for desktop GPU and not for mobile hardware

I don't think it's even possible to "optimize only for desktop GPU and not for mobile hardware".

I believe that essentially all of the optimizations "being done for desktop" would show comparable improvements when running on mobile hardware, even if the result is going from 20 fps to 40 rather than from 40 to 80.

If you are saying "the numbers are lower on mobile HW" that's probably true - mobile hardware tends to use lower end GPUs (fewer CUs, narrower/slower memory) and both CPU and GPU clocks are lower for power efficiency, but that's about it.

[finalzone]

amd dedicated desktop gpu, now are really good, open source driver are amazing.
it's a pity that mobile APU performance are really bad, seems that they are optimizing only for desktop GPU and not for mobile hardware

Mobile APU got more improvement notably on the hybrid laptop especially with more than 8GB RAM. Recent work done on Gnome Shell 3.22 under Wayland session allows better handling of APU. A10 7400P with R7 M265DX laptop runs smoother after upgrading llvm to 3.9 and kernel 4.9 without amdgpu support enabled (SI part of A10 7400P is unstable at the moment).

[Geopirate]

No clues? Really? Nvidia doesn't care one tiny little bit about specifications. They have a history of cheating on performance that goes back dozens of situations and decades of time. AMD's driver is very strict on adhering to standards, and as such it exposes bad design choices in OpenGL and many many bugs related to non-compliance in games and applications.

This is pure FUD. Nvidia has one driver with different shims for different platforms. This means the bulk of their engineers are working on the same code that works across all platforms. AMD has a lot of platform specific code and as a result a lot less engineers to work on this particular platform. This will change as they unify the code and more games switch to Vulkan. The future is looking really good for Vulkan as several big game engines are saying it's easier to work with than DX12.

[duby229]

This is pure FUD. Nvidia has one driver with different shims for different platforms. This means the bulk of their engineers are working on the same code that works across all platforms. AMD has a lot of platform specific code and as a result a lot less engineers to work on this particular platform. This will change as they unify the code and more games switch to Vulkan. The future is looking really good for Vulkan as several big game engines are saying it's easier to work with than DX12.

Yeah, nVidia most definitely leverages their cheating across all supported platforms. I can post links.... again... if you really want to them one more time.

[duby229]

Yeah, nVidia most definitely leverages their cheating across all supported platforms. I can post links.... again... if you really want to them one more time.

Pure FUD my ass. Pure 100% fact!

[duby229]

The company that has harmed linux gaming the most has been nVidia. First of all they have zero interest in standards compliance and as such their driver appropriately reflects there stance. Check ANY game developed on nvidia hardware and you can yourself verify that game is not standards compliant. What that effectively means is that almost no game ported to linux is OpenGL compliant, they ALL have have serious bugs that Nvidia not only condones but guarantees.

[duby229]

The company that has harmed linux gaming the most has been nVidia. First of all they have zero interest in standards compliance and as such their driver appropriately reflects there stance. Check ANY game developed on nvidia hardware and you can yourself verify that game is not standards compliant. What that effectively means is that almost no game ported to linux is OpenGL compliant, they ALL have have serious bugs that Nvidia not only condones but guarantees.

EDIT: And of course we already know the desire for compatibility is a bane. Compatibility requires duplicating behavior even if it is absolutely insane.

[johnc]

The company that has harmed linux gaming the most has been nVidia.

This board is a trainwreck.

[pal666]

Radeon RX 480 80.63
GeForce GTX 1060 80.75
Radeon R9 Fury 85.41
The NVIDIA Linux performance with their binary driver was meanwhile better with the GTX 1060 competing against the Radeon R9 Fury for this particular game

well, this explains why we are getting articles "i had to go with nvidia because i can't count"