Mesa 18.0/18.1/18.2 RadeonSI + RADV Benchmark Comparison With Radeon RX 580 / R9 Fury / RX Vega 64

[phoronix]

Phoronix: Mesa 18.0/18.1/18.2 RadeonSI + RADV Benchmark Comparison With Radeon RX 580 / R9 Fury / RX Vega 64

For those currently making use of Ubuntu 18.04 LTS with its default graphics stack (Linux 4.15 + Mesa 18.0) and are wondering if it makes sense upgrading to a newer version of the Linux kernel and/or Mesa, here is an extensive Mesa+AMDGPU comparison testing four graphics driver configurations across three popular AMD Radeon graphics cards.

http://www.phoronix.com/vr.php?view=26513

[M@GOid]

I'm getting regressions post Mesa 18.0, on a RX 570, running Kubuntu 18.04. For example, American Truck Simulator get some screen corruption, Deux Ex have the entire screen corruption with a black canvas taking over everything.

The last week Oibaf and Padoka PPAs won't even let me open Firefox or Chrome because some Wayland lib error (I'm running Xorg!). I only hope things to improve on the weeks ahead.

[prazola]

I'm getting regressions post Mesa 18.0, on a RX 570, running Kubuntu 18.04. For example, American Truck Simulator get some screen corruption, Deux Ex have the entire screen corruption with a black canvas taking over everything.

The last week Oibaf and Padoka PPAs won't even let me open Firefox or Chrome because some Wayland lib error (I'm running Xorg!). I only hope things to improve on the weeks ahead.

Regression on my r9 390x also, mad max, shadows of mordor, alien isolation, tomb raider runs at sub 60 fps at 1080p.

[fuzz]

The last week Oibaf and Padoka PPAs won't even let me open Firefox or Chrome because some Wayland lib error (I'm running Xorg!). I only hope things to improve on the weeks ahead.

I had this a few weeks ago too. I had to install a new package due to some library changes.

https://wordpress.padoka.org/2018/06/09/latest-mesa-breaks-wayland-workaround/

[M@GOid]

I had this a few weeks ago too. I had to install a new package due to some library changes.

https://wordpress.padoka.org/2018/06...nd-workaround/

Thanks for the tip.

[Pahanilmanlintu]

What about features, like freesync or wattman? Worth upgrading?

[Wielkie G]

The MadMax 18.2 results look very suspicious, as both 580, Fury and Vega 64 got the exact same framerate, down to 100th of a fps, for all of the tested locations. Same goes for 18.1 Vega64, which is the same as the 18.2 as well.

Could you please verify these?

[oooverclocker]

I'm getting regressions post Mesa 18.0, on a RX 570, running Kubuntu 18.04. For example, American Truck Simulator get some screen corruption, Deux Ex have the entire screen corruption with a black canvas taking over everything.

Unfortunately I can confirm corrupt rendering with an RX 580 8GiB in Deus Ex.
It was a wonderful year 2017 for RadeonSI but the performance improvements seem to slowly reach the limit with RadeonSI and Polaris and instead of spending a lot of effort to improve frame rates further I have to suggest putting a bit more significance to stability, smooth frametimes and compatibility. I don't know how often it's the case but I think when there are games that break stuff or that need fixes that make RadeonSI slower / delivering more inconsistent frame rates it might be more and more important to contact the developers instead of fixing it in RadeonSI - if that is the case for some regressions.

I like to thank Michael for this nice test but I am concerned that when we compare gaming performance the same way Windows-related testers do we might get the same FPS-targeted drivers that deliver a worse gaming experience. It would be fantastic when we could decide to measure only the frame time and how much it varies. We should honor those who make our video games more smooth and more enjoyable instead of creating wrong inducements with these FPS counts that don't show much about the smoothness. To be able to compete we need console-like performance and stability. Otherwise there is no reason to prefer Linux over Windows for gaming and Linux consoles will not be successful.
And finally Linux can't become successful on the desktop without these preconditions.

Regarding RADV I am impressed how much performance is still left behind but regarding the benchmark results it looks like development concentrates on improvements in specific games, correct me when I'm wrong. Of course one can do that and it might work fine but there is the danger that you improve one game and the others run worse, then you jump to another one which improvements make the remaining performances worse and so on.

So I would say that this time Michael grabbed a less optimal state of the project than the times before, unfortunately I can't post enthusiastic comments today and I hope a lot that people feel more encouraged to report regressions in time. And of course I'm looking forward to see that these issues have been taken care of in the next benchmark results.

[theriddick]

Ugh, seems the Vega64 wants to go faster but is being significantly held back.
Shame we don't see the huge jumps like the 980ti to 1080ti as with the Fury9 vs Vega64... a forever head scratcher of a card...

[lucrus]

To me those are great steps forward, especially because of my brand new RX 580! (used to have a RHD 6450 before...)