For instance, my GPU have low DRAM and it's on a laptop, I got a solid 60 fps Dark Souls III on Windows, whereas I face 50 ~ 30 fps on linux with micro stutters every now and then.
Besides that, my experience with Linux gaming is that it was absolutely not plug and play, it required me a lot of experimenting, changing proton versions, keeping myself up-to-date on the latest env variable to fix issues with my games, disabling composition on your DE, etc etc and etc.
I used to have fun tinkering with this stuff, nowadays I don't have the time/energy/will to educate myself on the latest tools to run a game, so I just keep a windows partition for gaming purposes.
- Use the latest Proton-GE or Proton Experimental (and switch to its bleeding edge branch). That fixes the anticheat and it includes patches to work around the shoddy DX12 logic in the game.
- If you're running on RDNA2, you might want to install mesa 22 or the latest git version. This is crucial for the best performance.
- On most AMD cards, it's worth trying to force the GPU frequencies to max. For some reason the GPU doesn't stay in high perf mode and it's a cause of frame drops.
The game stutters a bit on Windows and frame rate is inconsistent, while I have stable 60 fps on 4k Maximum quality on Linux. I've played 6h non stop today.
- Overwatch averages more stable framerates than equivalent hardware playing on Windows
- Apex Legends seems to run head-and-shoulders above what I experienced on Windows
- Half Life, Garry's Mod, and all of the Source engine games have wonderful native ports
- Older games like Diablo 2 are actually playable on Linux, where they're almost completely broken and dead on Windows
- This one is purely anecdotal, but MMOs seem to be much lighter on Linux. Games like Warframe, Diablo 3, and Final Fantasy 14 are light as a feather with accommodating hardware.
YMMV, but I'd say the state of gaming on Linux is very good right now. Maybe not as competitive as Windows is in market share and title compatibility, but it's far surpassing the selection (and performance) of PC games on Mac in it's glory days, and well outpacing the state of gaming on Mac now. If Valve continues to focus on their core competencies, we might finally have the gaming duopoly that we've rumored of for decades.
Truthfully, hacks might be needed, it's not grandma time yet for Linux gaming, but the hard part (Windows translation layer) is working really well.
The biggest obstacles (at least for Proton) keeping me from deleting my Windows partition:
- Online multiplayer games with anti-cheat (solved or soon to be solved by EAC + Proton integration)
- The fact that not every game "just works". I frequently run into games (most recently, Elden Ring) which don't even launch, and the error/bug reporting experience is terrible. I'm a relatively technical person, and yet even after going through the trouble to get logging set up and figure out what is even going on when the game doesn't launch, there is still often not a clear path to fixing my problem. Obviously some of this is just inherent in the combinatoric explosion of problematic cases you get from running a DX -> Vulkan compatibility shim for an OS with 100s of distros, but it's a big obstacle.
- I recently got a top of the line GPU, but before that the performance difference was occasionally enough that I would just switch to Windows for a game that otherwise worked OK on Linux.
There are tons of reports on Proton DB (and my experience agrees) where there are slow downs and stutters that just don't happen on Windows. Spyro: Reignited and Batman Arkham City are some recent examples for me.
I've played a few games that I can run at 4K with frame rates I don't notice any problem in, where with Proton it isn't the same experience.
For some older titles though I've been relatively impressed by Proton and e.g. CrossOver. Like Shadow of the Tomb Raider or Witcher 3. You can't play simple 2D games released this year but a 5 year old AAA game can run mostly OK, stuff like that.
There is also a flip side to the billion knobs thing which is that Linux might actually be a reasonable route to run some older games in the future, e.g. games like Deus Ex HR had some fun mods but some of them eventually became incompatible with newer Windows versions. It's not a guarantee (some of these games/mods do insane things) but for some classes of games, things like patched proton/wine builds may be a valid way to preserve them more easily into the future.
However, when something like ALVR works on Linux and SteamOS 3.0 releases universally I I'll at least make a partition for it and try to use it as my primary gaming OS.
(Flex ahead) Not had very many gaming related problems with my Steam Deck, so I am hopeful for the universal release.
But my personal gaming is dictated a lot by what my friends are playing, which means I prefer being ready to start a new game quickly without much hassle. And plenty of competitive multiplayer games are unsupported entirely on Linux.
I’m super excited for the future, but it just doesn’t fit my needs yet.
Games without anti cheat almost always work.
I use Linux for everything except unsupported gaming.
I've heard it might be playable "soon": https://old.reddit.com/r/linux_gaming/comments/t07q1g/shared...
https://appdb.winehq.org/objectManager.php?sClass=version&iI...
(Given you don't want to join any public servers due to the anti cheat not working yet)
Newer CPU's post 2008, have the elements of hardware drm Microsoft, intel, amd and the game and content industries have been testing. To get rid of local file access.
See here on the patent 20 years ago, aka windows 10/11 are the implementations of research that had been going on for 20 years to get rid of honest file access:
https://www.theregister.com/2001/12/13/the_microsoft_secure_...
Basically when you open a file on a normal PC you get honest hex values (instructions) that made cracking copyprotection on software easy because you have access to the raw CPU instructions. It's hard to make eletronics that don't allow piracy because all a file is, is electric charges in memory, and the computer industry knows this. AKA electricity is not scarce so supply can always meet demand trivially. So they needed to come up with bullshit to justify breaking CPU circuits without killing performance (aka "Security or drm"), has a negative impact on performance and there's also the worry of bugs bricking your system if you do it wrong at the hardware level... this is why it's taken them 23+ years since the advent of trusted computing research that began in the mid 90's to finally get rid of honest executables.
When you compile a program in c, or any other programming language for 99% of computing history you'd get honest instructions in the resulting executable and files for the program. The computer industry wants to change this for their clients, the movie, game, and music industry. So they can enforce copyright at the hardware level. This is hard without breaking a computer, and they don't want a repeat of the pentium fdiv bug that caused a massive recall.
https://en.wikipedia.org/wiki/Pentium_FDIV_bug
That is why mirosoft and intel/amd have been very very careful their trusted computing research and testing with certain big enterprise companies and consoles. Consoles is where trusted computing experiments in order to prevent game piracy.
All this to prevent you form getting honest files.
To give you an example of what's been goin on for 23 years...
Ultima 9 was cancelled (but then later restarted), AKA EA was planning on making Ultima 11, 12, 13, as local applications (you get complete files on your computer) WITH multiplayer embedded in the EXE, but the success of UO, changed the entire direction of PC Gaming. They wanted to kill piracy, this is the attitude of the game industry:
https://www.theregister.com/2012/08/22/ubisoft_says_over_90_...
Basic computer science 101, two or more computers networked together become and behave as a single computer. AKA every program can be divided into two or more subprograms and run over a network, that means you lose control of your PC if you take up any client-server exe. AKA there is no program that can't be turned into a client-server app, why would you want all your applications dependent on a 2nd computer in microsoft or other companies office when there is no rational reason for this? Since that means you no longer own your pc.
We already had infinity multiplayer games in the 90's in quake 2, go listen here at John carmacks comment.
(somewhat paraphrasing) "No limit to the number of players inside a multiplayer world"
https://youtu.be/TfeSMaztDVc?t=96
So no ultima online, everquest, guild wars 1 ad wow are literally just RPG's with network multiplayer removed and dumped into a seperate exe file and sold back to the public as if it were some new type of game. It isn't, it's the same game with fraudulently coded network code. They wanted to kill us getting complete games with multiplayer that were infinitely copyable like quake 2.
That is why modern Unreal engine FPS games like Transformers fall of cybertron have their multiplayer "shut down" because the general public bought into the MMO scam that led to steam. Steam was valves attack on game ownership and undermining local binaries. Note as steam grew, they started allowing devs to force update games to tie them to steams backend, aka many games won't work without steam loading, all pre-steam games work just fine minus "mmo's" which were the first drm games.
So yes the general gaming public were scammed and lead to the stolen PC game apocalypse we've been suffering for the last 23+ years.
So we have an unreal engine game (Transformers FoC) (2014)
And we got UT2004 a game from 2004 that has more features then modern incarnations, how did we end up going backwards in time in PC gaming?
They want to kill honest binaries (aka, access to raw hex instructions of programs allowing them to be cracked). UWP games on windows 10 was a trial run of trusted computing tech where windows update disabled cracked exe's, that's why UWP games can only run on certain verisons of windows 10. Windows 11 is that on steroids, Microsoft is lying because they know the average person using windows is clueless and tech illiterate. They learned that in 1997 with ultima online, everquest in 99 and wow in 2004. Client-server software is the same as buying a program with missing files. It's literally a broken application.
This is who they are designing it for, big media companies:
https://www.theregister.com/2001/12/13/the_microsoft_secure_...
What is trusted computing?
TC provides for a monitoring and reporting component to be mounted in future PCs. The preferred implementation in the first phase of TC emphasised the role of a `Fritz' chip - a smartcard chip or dongle soldered to the motherboard. The current version has five components - the Fritz chip, a `curtained memory' feature in the CPU, a security kernel in the operating system (the `Nexus' in Microsoft language), a security kernel in each TC application (the `NCA' in Microsoft-speak) and a back-end infrastructure of online security servers maintained by hardware and software vendors to tie the whole thing together.
The initial version of TC had Fritz supervising the boot process, so that the PC ended up in a predictable state, with known hardware and software. The current version has Fritz as a passive monitoring component that stores the hash of the machine state on start-up. This hash is computed using details of the hardware (audio card, video card etc) and the software (O/S, drivers, etc). If the machine ends up in the approved state, Fritz will make available to the operating system the cryptographic keys needed to decrypt TC applications and data. If it ends up in the wrong state, the hash will be wrong and Fritz won't release the right key. The machine may still be able to run non-TC apps and access non-TC data, but protected material will be unavailable.
The operating system security kernel (the `Nexus') bridges the gap between the Fritz chip and the application security components (the `NCAs'). It checks that the hardware components are on the TCG approved list, that the software components have been signed, and that none of them has a serial number that has been revoked. If there are significant changes to the PC's configuration, the machine must go online to be re-certified: the operating system manages this. The result is a PC booted into a known state with an approved combination of hardware and software (whose licences have not expired). Finally, the Nexus works together with new `curtained memory' features in the CPU to stop any TC app from reading or writing another TC app's data. These new features are called `Lagrande Technology' (LT) for the Intel CPUs and `TrustZone' for the ARM.
Once the machine is in an approved state, with a TC app loaded and shielded from interference by any other software, Fritz will certify this to third parties. For example, he will do an authentication protocol with Disney to prove that his machine is a suitable recipient of `Snow White'. This will mean certifying that the PC is currently running an authorised application program - MediaPlayer, DisneyPlayer, whatever - with its NCA properly loaded and shielded by curtained memory against debuggers or other tools that could be used to rip the content. The Disney server then sends encrypted data, with a key that Fritz will use to unseal it. Fritz makes the key available only to the authorised application and only so long as the environment remains `trustworthy'. For this purpose, `trustworthy' is defined by the security policy downloaded from a server under the control of the application owner. This means that Disney can decide to release its premium content only to a media player whose author agrees to enforce certain conditions. These might include restrictions on what hardware and software you use, or where in the world you're located. They can involve payment: Disney might insist, for example, that the application collect a dollar every time you view the movie. The application itself can be rented too. The possibilities seem to be limited only by the marketers' imagination.