Radeon Windows 10 vs. Linux RadeonSI/RADV Gaming Performance

[Michael]

Subscribed.

Thanks & welcome!

[Aiua]

Michael, about that Civilisation VI benchmark, isn't 8x MSAA at 4k resolution a bit too much? I use a 24" 1080p monitor and never go beyond 4X MSAA because I see no benefit on it. In some games I use even less because FXAA is enough. Maybe at less MSAA settings the game can become more playable in Linux?

To be honest, while the comparison between Windows and Linux for Civ6 is a bit saddening, I find the game playable at 20FPS, it's not like a lot is going on graphically at any given time.

Anyway, overall I pleasantly surprised by the mesa results, especially after seeing the nvidia ones, I was expecting worse. Good job to everyone involved!

[Aiua]

When comparing the results (not reading the text at the same time), it is often difficult to figure out which Windows driver is being used (OGL, DX11 or DX12). In some cases, for example Tomb Raider and Shadow of Mordor, I didn't find that info at all.

When nothing is written for Windows, I think you can safely assume DX11

[labyrinth153]

Michael, about that Civilisation VI benchmark, isn't 8x MSAA at 4k resolution a bit too much?

I think the benchmark modes are set up this way, but I could be wrong.

[shmerl]

Can anyone clarify please. Is lower performance caused by lacking AZDO extensions, and games using less performant paths, or by already implemented extensions not being fast enough?

[dungeon]

Can anyone clarify please. Is lower performance caused by lacking AZDO extensions, and games using less performant paths, or by already implemented extensions not being fast enough?

Everything of these, plus OS differences... driver differences. With performance one never know, you just do benchmarking at random point of time to see where you are, but...

But there are now already newer driver for Windows, also The Talos Principle get new update yesterday:

http://steamcommunity.com/games/2575...00146533179334

These are things Michael didn't tested, as future always eating past... hope i clearfied something

[theriddick]

Not as bad as I was expecting, shame Civ/SOM/Talos lag behind so much. Interesting to see the AMD hardware doing better with the Unigine benchmarks compared to NVIDIA if I remember correctly...

I feel if AMD merged some of the optimizations from AMDGPU-PRO into their open driver then some of these odd ball games that chug would perform significantly better, I do remember Talos running ALLOT better on the PRO driver then open for example!

Micheal while you may not get a great deal of subscriptions, surely you get some decent ad-view revenue from these? (from other sites that reference here)

[Creak]

I wonder if there are softwares that can test the GPU and are well ported on both platforms (either app or games or engines...). I guess GPUtest is one of them, but are there more?

Effectively, since the perf differences between the games are also due to the quality of the Linux port, these number are good for the end-users, but not really interesting for the mesa developers. It's hard to know if the problems come from the port or the drivers.

Michael If we figure out a list of softwares that we know should be as good on Windows than on Linux, do you think you could manage to make an article around it?

[bug77]

That's a better showing than I would have guessed, but still, overall too inconsistent for my liking.

[Michael]

Michael If we figure out a list of softwares that we know should be as good on Windows than on Linux, do you think you could manage to make an article around it?

Only if the software is all fully-automated for PTS and/or there is a bunch of new subscribers.... so far this round of Windows 10 vs. Linux testing has proven to be a waste with minimal donations/new-subscribers not offsetting all the time spent on this round of testing, rather disappointed.