Valve updates Proton to 7.0, enhancing Steam Deck game compatibility

Daniel Sims

Posts: 1,376   +43
Staff
Why it matters: Valve has released a significant update for the Steam Deck's compatibility layer for Windows games. Proton 7.0 adds compatibility for more titles, support for Easy Anti-Cheat (used by Fortnite, Apex Legends, Rocket Arena, and many others), and fixes issues found in some games.

The latest update to the Windows-Linux compatibility layer Proton brings it to version 7.0. The new features and compatibility improvements it brings to Valve’s SteamOS arrive less than two weeks before the Steam Deck's official launch. Wine 7.0 and DXVK 1.9.4 are part of the patch as well.

Games newly playable through Proton include Anno 1404, Forza Horizon 5, Monster Hunter Rise, Disgaea 4 Complete+, Persona 4 Golden, Oceanhorn, Resident Evil 0, Wargroove, Yakuza 4 Remastered, and more. The update also fixes voice chat in Sea of Thieves; improves audio in Skyrim, Fallout 4, and Mass Effect; and adds controller support for games running through EA’s Origin client.

Other improvements include support for locally decoding H.264 videos and Easy Anti-Cheat (EAC) for games with Linux modules. Concerns about EAC support on the vast array of Linux kernels are the main reason Epic Games boss Tim Sweeney says Fortnite won’t support the Steam Deck.

Wine 7.0, which was rolled out last month, includes improvements to how it handles DirectX, Vulkan, 64-bit Windows-on-Windows, and a lot more.

Permalink to story.

 
It's going to be interested to see if Sweeny concedes and tries to put Epic Store on the Steam Deck now that Valve will go after Fortnite as fully playable OR if he's going to double down and start an arms race to patch anti-cheat and break compatibility.

I actually think that while Sweeny and Epic themselves might not go there, I am almost sure other DRM companies (Because really, that's what most 'anti-cheat' companies are at the end of the day: An extra step we wouldn't need if we could control online gaming with private servers like it used to be) will continue to fight Linux adoption.
 
I don't think Sweeny have any issues with putting their games on linux when possible, EAC is part of Epic and they actually made it possible to use in there. Sweeny said Fortnite wont be on linux for now because of possible workarounds on EAC on Kernel level which I think is understendable, and it is way too early to ask them to put their store on Linux as they still trying to establish their position on PC. I havent heard GoG or Ubisoft going full linux either.

EAC is more than a DRM, othervise Denuovo would be enough. All the wall hacks, transparent textures, autoaim bots are client-side, not server-side. If there is a competitive, global server then it is important any cheaters would be declined access to it. DRM is usualy used to single player content only (or digital media (+stream) etc).
 
I havent heard GoG or Ubisoft going full linux either.
Ubisoft did go on record to say if the Steam Deck does well, they'll consider putting their games back on Steam. Not heard anything about GoG though.

I'm looking forward for the NDA's to lift so we can see what the software is like as I wonder how it currently deals with EA's Origin when you load up an EA game in Steam.
 
Ubisoft did go on record to say if the Steam Deck does well, they'll consider putting their games back on Steam. Not heard anything about GoG though.

I'm looking forward for the NDA's to lift so we can see what the software is like as I wonder how it currently deals with EA's Origin when you load up an EA game in Steam.

Honestly I'm more interested in how this device does once windows is installed. Largely for emulator performance, as some of the Linux versions don't run as well as the windows version. The Steam Deck is starting to look like a better deal than the Odin Pro. Too bad it doesn't have dual SD card slot.
 
2 weeks from launch and Valve is still hammering away. This isn't good.
What isn't good? If you're referring to the fact they're working on proton still, that's normal, expected and was always communicated as an ongoing development.

The PS5 launched with an NVMe expansion port we couldn't use for year until they released a software update. At least the Decks SD expansion is working from launch.
 
Last edited:
Personally I think this is a wasted opportunity to bring PC gaming to ARM. We could have had smaller and more efficient hardware running our steam libraries. Instead we’ve got this enormous box that weighs considerably more than any handheld device I can think of and lasts 2 hours on battery. They are using a compatibility layer so they could have done that to get games working on ARM for those that don’t have native support.
 
Personally I think this is a wasted opportunity to bring PC gaming to ARM. We could have had smaller and more efficient hardware running our steam libraries. Instead we’ve got this enormous box that weighs considerably more than any handheld device I can think of and lasts 2 hours on battery. They are using a compatibility layer so they could have done that to get games working on ARM for those that don’t have native support.
Except that there is no non-Apple silicon ARM chip that does x86 emulation through dedicated hardware. And turns out that even the M1 knows how to drain a battery fast when gaming.
So since Valve doesn't employ CPU architects they'd have to use something like the Snapdragon 8 gen 1.
That thing can pull 11.2 watt, not far off from the 15 watt APU it has now. Being arm it would have to emulate every single game as they're all x86 which then would absolutely destroy performance, you'd have games running at 10fps if even that.
Just because the M1 runs x86 darn well doesn't mean other arm chips do so.

So, missed opportunity? I think not. A large part of why it's big and bulky seems to be due to valve just choosing to do so. Partly for the wealth of inputs and them still being easy to use. Partly to be somewhat repairable and cheap to construct. The Aya neo is considerably smaller.
Seeing the massive leaps in power efficiency the AMD 6000 APUs made a steam deck V2 could have significantly better battery life. Hell imagine those energy efficiency gains in combination with TSMC's 5nm process.
 
Except that there is no non-Apple silicon ARM chip that does x86 emulation through dedicated hardware. And turns out that even the M1 knows how to drain a battery fast when gaming.
So since Valve doesn't employ CPU architects they'd have to use something like the Snapdragon 8 gen 1.
That thing can pull 11.2 watt, not far off from the 15 watt APU it has now. Being arm it would have to emulate every single game as they're all x86 which then would absolutely destroy performance, you'd have games running at 10fps if even that.
Just because the M1 runs x86 darn well doesn't mean other arm chips do so.

So, missed opportunity? I think not. A large part of why it's big and bulky seems to be due to valve just choosing to do so. Partly for the wealth of inputs and them still being easy to use. Partly to be somewhat repairable and cheap to construct. The Aya neo is considerably smaller.
Seeing the massive leaps in power efficiency the AMD 6000 APUs made a steam deck V2 could have significantly better battery life. Hell imagine those energy efficiency gains in combination with TSMC's 5nm process.
No, you’re wrong it is a big missed opportunity. I don’t really know why you would assume Apple should make the part, you can’t buy Apple chips from Apple and they aren’t designed for gaming. I would have thought one of Nvidias Tegra like solutions would be better. It’s definitely out there. The first company to crack ARM gaming will make a lot of money. X86 is dying, we will not be using chunky 2 hour battery handheld devices forever. If it won’t be Valve who does it, it will be someone else. Don’t forget the switch is an ARM device. ARM gaming isn’t a new thing.

Also there is no massive leap in efficiency with the Ryzen 6000 series. It’s a tiny increase. It’s still X86, the chips even use the same power just clock slightly higher for doing so, calling it “massive” is laughably inaccurate.
 
Also there is no massive leap in efficiency with the Ryzen 6000 series. It’s a tiny increase. It’s still X86, the chips even use the same power just clock slightly higher for doing so, calling it “massive” is laughably inaccurate.
Wow, you really don't do your research before posting do you? I don't really know what I expected in all fairness, every post you make seems to be the same.
 
No, you’re wrong it is a big missed opportunity. I don’t really know why you would assume Apple should make the part, you can’t buy Apple chips from Apple and they aren’t designed for gaming. I would have thought one of Nvidias Tegra like solutions would be better. It’s definitely out there. The first company to crack ARM gaming will make a lot of money. X86 is dying, we will not be using chunky 2 hour battery handheld devices forever. If it won’t be Valve who does it, it will be someone else. Don’t forget the switch is an ARM device. ARM gaming isn’t a new thing.

Also there is no massive leap in efficiency with the Ryzen 6000 series. It’s a tiny increase. It’s still X86, the chips even use the same power just clock slightly higher for doing so, calling it “massive” is laughably inaccurate.
I'm mentioning Apple because they're the only ones that have proven they can emulate x86 on ARM with good (extremely good even) performance. Everyone else has to force it in software, taking the most performant mobile arm chips possibly available during the steamdecks development such as the aforementioned Snapdragon 8 gen 1 it would remain to be seen if they got the CPU grunt. Not to mention the overhead this emulation overhead adds kinda negates ARM being more efficient.

No chance in hell that all games will suddenly come compiled and optimized for ARM. Tens of thousands of games spread over thousands of developers, many likely not even around anymore because it also spans decades.

So to sum it up.
* Performance might not be good enough, it would definitely be more random with which games run well and which don't.
* The chip they'd have to use isn't even that much better when it comes to power draw.
* Even more headaches when it comes to titles with anti-cheat software (in addition to having a Linux platform).

Or in short, why bother. Sticking to x86 simplifies things a lot and trying to launch a Linux based handheld capable of running almost the entire steam library is high enough a bar to aim for. For a next iteration if arm is a substantial step forward they could consider it then.

I don't see x86 going away in favour of arm soon though. Heck if anything my money would be on RISC-V replacing them both.
 
Relax on the x86 emulation, guys. This is a first generation product and we don't even know how well it will sell. The future of ARM is also still up in the air.

If it does well and they get the kinks out of the Linux compatibility layer then we can start talking about developing chips specific to the platform.

Right now we're at the "lets bring a product to market for the absolute lowest cost we can and see how it sells" phase.

I also don't see the 2 hour battery life being too big an issue. If I'm on a flight I usually have a large battery bank with me. I don't want the steam deck as a daily driver, I want it so I can chuck it in my bag and have it in my hotel when I travel.
 
Last edited:
I also don't see the 2 hour battery life being too big an issue. If I'm on a flight I usually have a large battery bank with me. I don't want the steam deck as a daily driver, I want it so I can chuck it in my bag and have it in my hotel when I travel.
That's also worst case. The official spec said 2-8 hours. So a lot of games might get something like 4.5 hours which is what the switch states as the minimal battery life (although in practice that can be about a hour less than that as well). Basically if you need the battery to last you could stick to simpler games or turn the graphics settings a notch
 
That's also worst case. The official spec said 2-8 hours. So a lot of games might get something like 4.5 hours which is what the switch states as the minimal battery life (although in practice that can be about a hour less than that as well). Basically if you need the battery to last you could stick to simpler games or turn the graphics settings a notch
I see it the same way I saw it with my switch, I got uncomfortable after about an hour and put it down. It wasn't any 1 thing, either. It was my hands got tired of holding it, my eyes got tired of looking at the screen, my seating positions became uncomfortable. I see my self putting it down due to ergonomics long before I reach the limits of the battery.
 
Personally I think this is a wasted opportunity to bring PC gaming to ARM. We could have had smaller and more efficient hardware running our steam libraries. Instead we’ve got this enormous box that weighs considerably more than any handheld device I can think of and lasts 2 hours on battery. They are using a compatibility layer so they could have done that to get games working on ARM for those that don’t have native support.
As we see with upcoming amd 6000 family, there is no advantage on efficiency apart of node process. Steam Deck 2 with new x86 compatible portable amd apu will be even better amd I see no advantages for anything, which will have to emulate instructions in order to run modern games.
ARM is a good system, but so is x86. On the same node level energy consumption will be on pair, more or less. Check the article here on amd 6000.
 
I never knew that Steam Deck would receive such an overwhelming response from the audience. But, I think its quite early to say about its success.
 
Back