• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Linux gaming thread, what games are you currently playing?

AVATARAT said:
On the second and third runs, everything works normally because the data is cached and performance is better; on the first run, the access time is too long.
It’s normal for the FPS to vary a few points.
Click to expand...
I've experienced the caching regarding benchmarking, you may gain a few more fps/points at most, these results are varying wildly. Looking at max fps alone, there's a difference of 83 fps between the worst and the best runs, and those runs weren't the first one's performed.

AVATARAT said:
I can run both version, DX and OGL.
Click to expand...

I can run DX and OGL under vanilla Wine, running any form of Proton limits me to DX only, with OGL producing a black screen. But running the native build of the benchmark, I can run OGL.
 
Mazzspeed said:
I've experienced the caching regarding benchmarking, you may gain a few more fps/points at most, these results are varying wildly. Looking at max fps alone, there's a difference of 83 fps between the worst and the best runs, and those runs weren't the first one's performed.
Click to expand...
Max and Min are simply recorded as numbers; they can change for a variety of reasons—yes, it would be nice if they were more stable, but they aren’t. The average value and the results, however, are usually similar, except when something is using or freeing up CPU resources.
Sometimes, damaged or poorly functioning libraries can reduce performance in both Windows and Linux.
Mazzspeed said:
I can run DX and OGL under vanilla Wine, running any form of Proton limits me to DX only, with OGL producing a black screen. But running the native build of the benchmark, I can run OGL.
Click to expand...
Hum, yes, you are right.

Running Unigine Valley under XWayland often resolves black screen issues on Wayland because it uses the familiar X11 OpenGL context instead of native Wayland EGL.
 
AVATARAT said:
Max and Min are simply recorded as numbers; they can change for a variety of reasons—yes, it would be nice if they were more stable, but they aren’t. The average value and the results, however, are usually similar, except when something is using or freeing up CPU resources.
Sometimes, damaged or poorly functioning libraries can reduce performance in both Windows and Linux.
Click to expand...

The minimum value can vary depending on background tasks affecting the loading of textures between scenes (which I noticed influenced the minimum result under Unigine Heaven quite substantially). But I've never seen such a massive variance in the max fps before, and as stated even with gamemode running and power management set to performance, Unigine Heaven as a benchmark was struggling to max out 1C/1T of my 14C/24T i9-10940x - Something Unigine Superposition is able to do, even if it is 1C/1T.

Both benchmarks are very old now, released when 4C/8T CPU's were used by most, and games were mostly single threaded; they're not really a good representation of relative performance on a system with 8C/16T and higher - No bench should vary that wildly between runs on the same system with no additional tweaks made. I know for a fact that I've got no damaged libraries here...
 
Mazzspeed said:
The minimum value can vary depending on background tasks affecting the loading of textures between scenes (which I noticed influenced the minimum result under Unigine Heaven quite substantially). But I've never seen such a massive variance in the max fps before, and as stated even with gamemode running and power management set to performance, Unigine Heaven as a benchmark was struggling to max out 1C/1T of my 14C/24T i9-10940x - Something Unigine Superposition is able to do, even if it is 1C/1T.

Both benchmarks are very old now, released when 4C/8T CPU's were used by most, and games were mostly single threaded; they're not really a good representation of relative performance on a system with 8C/16T and higher - No bench should vary that wildly between runs on the same system with no additional tweaks made. I know for a fact that I've got no damaged libraries here...
Click to expand...
Unigine Heaven depend on Single thread but far less than Unigine Valley, yes they are very old but they have their use case. As I said up, Valley if very good for Single thread benchmarks.
Unigine Superposition is pure GPU benchmark, you will get similar results with different CPUs and same GPU.
I’ve been using them and other programs for years, and they work quite consistently. If something interferes with them, the performance usually drops, which is a plus because it makes it easy to see when there’s a problem. Ultimately, these programs are tools for identifying issues. In Windows I normally check VC++/DX or system files.
 
AVATARAT said:
Unigine Heaven depend on Single thread but far less than Unigine Valley, yes they are very old but they have their use case.
Click to expand...

As a benchmark, Heaven is not CPU limited on a modern(ish) CPU as it can't actually max out 1C/1T. Watching htop, it's not even pushing a single core/thread to full boost clocks even with gamemode enabled and power management set to performance for good measure - it just jumps cores all over the place, and at 1080p it's certainly not GPU limited - Therefore, as a means of benchmarking modern(ish) systems, it's quite useless and provides questionable results that vary 'wildly'. Superposition on the other hand 'can' actually max out 1C/1T on a modern(ish) CPU, and provides results that are reproducible and generally within the ballpark of one another.

I have to emphasize once again, the wildly varying Heaven results I've posted aren't the result of any issue with my system, my PC is running perfectly. Heaven is simply up there with other benchmarking hero's of a bygone era like 3D Mark 2001SE and too old now to really hold any relevance.
 
Currently playing Crimson Desert. Upon first load the Logo screen and skill video popups were just color bars but I updated Cachy Proton to the newest version which seems to have fixed it. There is a strange blockiness to textures/geometry where it kind of looks like the corners of a boundary box are showing up. Performance is good; getting 70-90fps at 3440x1440 and Ultra settings (Native res - No FSR). I did get one crash about 2 1/2 hours in but the system didn't lock up so I was able to just kill/restart it..
 
When have some time to dedicate (30+ min):
- The Division 2
- Titanfall 2
- Halo (Currently CE)

In that order

When I don't know if I have 5 minutes or what:
- Goat Simulator 3
- Orcs Must Die 3
- Beam.NG Drive

Also in that order.

Screenshot_20260320-164413.png
 
And a few more tests—yes, the result is a little lower, but I think there’s also an effect missing.
Best result was with Proton_slr_v3

1774140698514.png

1774140717567.png


CatZilla_720p_148590_W.png

CatZilla_2k_87887_W.png
 
Mazzspeed said:
Is there a point to these benchmarks, can you provide some evidence there's an effect missing?
Click to expand...
I just did it because it's fun :)

Hmm, after the restart, everything seems to be working fine, strange. I mean the picture is now same in both OS.
 
  • Like
Reactions: ChadD
like this
AVATARAT said:
I just did it because it's fun :)

Hmm, after the restart, everything seems to be working fine, strange. I mean the picture is now same in both OS.
Click to expand...

How did you even install this game under Linux? I've tried Lutris, the installer runs but the game won't launch, and the software appears under Steam but seems have gone unsupported for so long that you can no longer download it.
 
Hmmm. Your 720p result looks far lower than mine. I didn't notice any missing effects. PC specs in sig:

Catzilla Results 720p.png
 
Last edited:
Second run at 720p, can't run any other test unless I pay for it (there's no way I'm paying for this).

Catzilla 720p run 2.png
 
Last edited:
Mazzspeed said:
I don't know where to put the key, and I can't seem to create an account as the Captcha won't verify that I'm human! This benchmark is largely abandonware.
Click to expand...
The Upgrade part of the app. Works fine after pasting & clicking send.

1440p results - specs in signature
1774154793388.png

I'm really not sure why its rated so low... probably the difference between the 7900 XT & the RTX 5090
 
CrimsonKnight13 said:
The Upgrade part of the app. Works fine after pasting & clicking send.
Click to expand...

Daymn, how did I miss that! The scores on this bench seem...Odd somehow? I'm running the bench off a mechanical HDD, would probably get slightly higher results off my SSD...

1440p results, specs in sig:
Catzilla 1440p latest.png
 
Last edited:
Mazzspeed said:
Hmmm. Your 720p result looks far lower than mine. I didn't notice any missing effects. PC specs in sig:

View attachment 792908
Click to expand...
The bench cannot display more than 5 numbers. My result is missing the last number:D
Check out these results:

1774178508776.png


I am not sure why but in Linux If I want to change resolution, I need to restart the Launcher after each run.
 
Bankie said:
Currently playing Crimson Desert. Upon first load the Logo screen and skill video popups were just color bars but I updated Cachy Proton to the newest version which seems to have fixed it. There is a strange blockiness to textures/geometry where it kind of looks like the corners of a boundary box are showing up. Performance is good; getting 70-90fps at 3440x1440 and Ultra settings (Native res - No FSR). I did get one crash about 2 1/2 hours in but the system didn't lock up so I was able to just kill/restart it..
Click to expand...
Seen the boundary box stuff in windows reviews, seems to be the game and not a proton thing.
 
AVATARAT said:
The bench cannot display more than 5 numbers. My result is missing the last number:D
Check out these results:
Click to expand...

WTF? That's pretty crappy. Most of these benchmarks are literally abandonware at this point.

AVATARAT said:
I am not sure why but in Linux If I want to change resolution, I need to restart the Launcher after each run.
Click to expand...

I definitely don't have to do that.
 
64bit Steam's been released as beta. I've been running it for a few days now and it hasn't missed a beat, runs perfectly with hardware acceleration enabled under Nvidia hardware now, even Big Picture Mode works great. 👍
 
I thought it was released already? The steam-native package has been removed from arch for a while now.
 
Nobu said:
I thought it was released already? The steam-native package has been removed from arch for a while now.
Click to expand...
It's been running in a 32-bit container for a while, native or not. Now they're switching over to a 64-bit version instead, which will allow the Steam Interface & Input APIs to work properly with Wayland.
 
Crazy that Linux is up 3.10% but the top 6 distros there only make up about .5% of it.
 
Playing Crimson Desert on Bazzite. I've got a grip on how to do Proton version switching if needed, which it rarely is outside of bleeding edge new games. It's actually extremely easy.

Thr game runs fine with the current Proton versions from steam but with limited FSR only until the bleeding edge updates make it through the pipeline, But I wanted FSR 4.1 right now so I slapped on Proton from CachyOS and a couple of commands for startup in Steam and I've got FSR 4.1, Frame Gen 2x, and HDR all working. Ray Regeneration is also working fine but just like on Windows the perf hit in this game is too high. It looks stunning and runs great at 4k with VRR.

I used the same options to get FSR 4.1 working in ARC Raiders. Looks and runs fantastic. So much sharper.
 
Bankie said:
Crazy that Linux is up 3.10% but the top 6 distros there only make up about .5% of it.
Click to expand...
If you expand the survey to show Linux only and click on Linux version to see the breakdown of distro used, things make slightly more sense. Although, it seems CachyOS is now all lumped into the '0 64 bit' category, with '64 bit' possibly representing the new Steam 64 bit Beta and Fedora/PopOS/etc lumped onto 'Other'.

Steam breakdown March 2026.png
 
Been playing ME3 Legendary Edition. It has mostly been smooth except for about three crashes but only during conversations. I blame it on having the EA version of the game. Probably found it on a really good sale a few years ago through EA's store and bought it there. Still regretting that purchase. I'll probably end up buying it through Steam during a really good sale at some point just to avoid having to use whatever EA is calling their store/launcher. Will still have to deal with the EA launcher stub through Steam but that's usually not an issue.
 
The open source AMD graphics drivers in the Linux kernel (and used by FreeBSD) are largely submitted by AMD engineers.
 
uOpt said:
The open source AMD graphics drivers in the Linux kernel (and used by FreeBSD) are largely submitted by AMD engineers.
Click to expand...
I don't believe that's the case anymore. I believe they're mostly developed by Valve, and that's not a bad thing.
 
You must log in or register to reply here.
Back
Top