Okay, after some fiddling around I managed to get up-to-date Catalyst on this system. However, the impact was quite minimal and most of the time non-existent. I've added corrections to the article in places that need them and updated one of the benchmark graphs where the update yielded a performance increase. Games like Shadow of Mordor and benchmarks like Heaven were unaffected by the driver update and after the update HL2: Lost Coast started running without graphical glitches but the performance remained the same.

View this comment - View article - View full comments

I verified the FGLRX version and it's indeed the old driver, not 15.7. That kinda makes the results pointless in that it's not the most recent driver version. However, it appears to be nearly impossible for me to install the 15.7 driver so that it works. I had no luck with the installer and the .deb they provided simply boots me into TTY. I'm not really determined enough to try to figure out what is breaking and where, so I'll just wait for someone to properly package that driver. Maybe the Ubuntu Graphics PPA team will figure something out in near future. If I do get the drivers up and running I'll update the results.

View this comment - View article - View full comments

Quoting: GuestVery interesting read, especially the Gallium Nine part.

Quoting: BTREYou should try benchmarking Talos.

Yeah, it's the second (after Unigine) "real" benchmark that is free (available in demo) for everyone. For some reason you will never see TTP on Phoronix...

It can provide some interesting stats too http://steamcommunity.com/app/330710/discussions/0/620696522118520223/?l=polish [External Link]

PS why FGLRX is slower without tessellation?

I'll see what I can do with Talos Principle. Also, when you look at the graphs make sure to see what the framerates actually are. Even though at a quick glance it might look like FGLRX is slower without tessellation, this is not actually the case. In fact, FGLRX ran about 64% faster without tessellation (77.6 FPS vs 47.2 FPS).

View this comment - View article - View full comments

Quoting: sorathAt first thanks for benchmarks for AMD videocard.

Could you clarify please: FGLRX 15.20 is Catalyst 15.7? On Ubuntu Catalyst 15.3beta AFAIR also was marked as fglrx 15.20.

Could you also compare game's performance between Linux and Windows ports with the same video cards if WIndows OS is available for you?

I'm not entirely sure if the FGLRX I installed from the repositories was 15.7 but I think it was. I still don't understand the Catalyst version numbering too well.

As for the Windows comparisons, I will not run such things on my hardware. The only Windows license I had was thrown to Liam and getting a new license is too expensive. In addition to that I don't like Windows a whole lot and I don't want to put up with its quirks for the sake of booting it up maybe 4 times a year for some comparisons. Also, I would have to get a new hard drive because I'm not wasting a gigabyte of my current storage for something like Windows.

View this comment - View article - View full comments

Quoting: xyzzos"Open driver" does not imply subpar (performance) or anything of the sort. If
anything strongly favours subpar results is working without proper hardware
documentation and with little or no financial support. *Not* being free/open.

You are right, being open doesn't mean that it's inferior. However, context is very important here. I haven't seen an open GPU driver that outperforms a proprietary driver and I come from Nvidia background where the performance has been particularly crippled on open source driver so when an open source driver does perform well I'm quite excited. I'd be rather silly if I thought that open source software is inferior (and yes, I do often say "open source" instead of "free software" ) since my work is exclusively done using open source software.

View this comment - View article - View full comments

Quoting: melkemindSaying it's "completely wrong" almost implies they lied about it. More than likely, you and Liam must have some configuration settings that make the game run better. It would be helpful for all of us to know what those settings are (on your desktop, drivers and game). Perhaps you can help those of us who are getting numbers similar to Phoronix.

That's easy. Go to the Nvidia X Server Settings -> OpenGL Settings and disable Sync to VBlank. It was noted in the article that a recent driver update managed to turn that on without asking the user and regardless of previous settings.

View this comment - View article - View full comments

Quoting: ungutknutFor everyone interested in Windows vs. Linux Benchmarks of this game (don't klick if you're not prepared for bad news):
http://www.phoronix.com/scan.php?page=article&item=mordor-win10-linux&num=1 [External Link]

Here's another test coming to the same result:
https://www.youtube.com/watch?v=89G9qHrjS4A [External Link]

Seems SOM needs some heavy optimization work.

Those Phoronix benchmarks are completely wrong and that's one of the reasons why we ran our own benchmarks. You can see Liam's 970 beat a 980 Ti so whatever Larabel did was completely wrong. The Windows results might be accurate but Linux most definitely isn't. Take the Phoronix results with a boatload of salt.

View this comment - View article - View full comments

Quoting: OZSeafordWhat?????

I am on a GTX760, and I get an average of around 25-30fps on medium detail.

Is it because I am using the default driver? I am on version 340.76...

You should upgrade to 352. The requirements on the store page state that it's the only supported driver. I recommend the following PPA for driver updates: https://launchpad.net/~xorg-edgers/+archive/ubuntu/ppa [External Link]

View this comment - View article - View full comments

I'm just going to chime in and say DON'T FRICKING PRE-ORDER VIDEO GAMES OR ANYTHING ELSE FOR THAT MATTER. Thank you.

View this comment - View article - View full comments

Xorg-Edgers is a good PPA. All their Nvidia blobs have worked nicely.

View this comment - View article - View full comments