Valve's own operating system "SteamOS", based on a specific Linux distribution called Arch Linux, alongside their own Wine fork titled "Proton", have both opened up the doors for modern and AAA gaming on Linux machines, with their OS being developed specially for their Steam Deck handheld device. Since the release of the Steam Deck and SteamOS, many have wondered if the company could consider releasing their OS for general use as a distributable Linux distro to be installed in other devices and computers outside the Steam Deck.

That could all be a reality very soon, as Valve's designer Lawrence Yang has confirmed in an interview with The Verge that they are planning on bringing SteamOS support for other handheld devices in the market, more specifically, the ROG Ally.

Valve announced a few days ago a new update to the Beta channel of SteamOS, with version 3.6.9 adding a very specific general change which namedropped the ROG ally with the message "Added support for extra ROG Ally keys". While it was possible that Valve was simply adding support for the Steam Client on Windows to support these keys (given that the ROG Ally runs under Windows), The Verge asked directly if this was simply a Big Picture mode change, or if Valve is indeed planning to bring SteamOS to other systems:

The note about ROG Ally keys is related to third-party device support for SteamOS.
The team is continuing to work on adding support for additional handhelds on SteamOS.

Lawrence Yang mentions that work on bringing SteamOS to other devices, and even PCs, is currently "making steady progress", but their work is not ready quite yet, so it could still take some time to see SteamOS in other powerful handhelds and main PCs.

The Verge also asked about the possibility of dual-booting Windows on a Steam Deck, to which Yang replied:

As for Windows, we’re preparing to make the remaining Windows drivers for Steam Deck OLED available (you might have seen that we are prepping firmware for the Bluetooth driver). There’s no update on the timing for dual boot support—it’s still a priority, but we haven’t been able to get to it just yet.

Even though it could take some time to see SteamOS in other systems, the future for Valve's SteamOS, and Linux gaming as a whole, is looking better by the day, with more and more users using Linux and Steam for their gaming needs as days go by.

Source

 

[Xzi]

By PC I guess you mean x86 in general?
I agree x86 REALLY needs to die...

Lol what? So all console and PC gaming needs to die, leaving us with just predatory P2W and gacha garbage? I'd much rather see ARM gaming die off, save for emulation.

I'm scared Valve will never develop a "Steam Deck 2".

They will, but only when there are big hardware improvements available which won't cost a whole lot more to include. The appeal of Steam Deck is being several tiers above Nintendo's offerings in terms of performance, but only being a half tier or so above them in pricing. They'll leave the $800+ portable PCs/gaming laptops to other manufacturers.

Basically, expect Steam Deck 2 to be revealed after Switch 2 has been on the market for 3-4 years.

 

[LoggerMan]

By PC I guess you mean x86 in general?
I agree x86 REALLY needs to die...

Yes. If Apple can write code that lets most x86 apps run basically flawlessly on Mac, like they did when switched from PowerPC to Intel, then it must be possible to get x86 games to run well on modern ARM chips. There's some monster power in these chips. I guess it's inevitable as Windows is transitioning to ARM too now. One day playing on a old fashioned PC will be the niche way to play games, the way weird grandpa chooses to play it while everyone else is using ARM devices to play almost all old and new games.

 

[wiindsurf]

Lol what? So all console and PC gaming needs to die, leaving us with just predatory P2W and gacha garbage? I'd much rather see ARM gaming die off, save for emulation.

I mean PCs and consoles should transition to ARM, just like the Macs did, as well as the Nintendo Switch... I am also not a fan of mobile gaming.
x86 is an almost 50 years old Frankenstein and is a hinderance to mankind at this stage... x86 is older than the NES, about as old as the Atari, and it is the only component on a current PC that hasn't seen a true clean sheet modern redesign.

 

[Xzi]

I mean PCs and consoles should transition to ARM, just like the Macs did, as well as the Nintendo Switch... I am also not a fan of mobile gaming.
x86 is an almost 50 years old Frankenstein and is a hinderance to mankind at this stage... x86 is older than the NES, about as old as the Atari, and it is the only component on a current PC that hasn't seen a true clean sheet modern redesign.

Macs are the absolute worst platform for gaming for a reason, and Switch is severely lacking in processing power. ARM chips are great for lower power consumption, but they're still playing catch-up with x86 CPUs in every other aspect, otherwise that transition would've happened already. Die shrinkage is what will determine the limit for x86 performance, and we're still nowhere close to hitting that limit.

 

[subcon959]

I mean PCs and consoles should transition to ARM, just like the Macs did, as well as the Nintendo Switch... I am also not a fan of mobile gaming.
x86 is an almost 50 years old Frankenstein and is a hinderance to mankind at this stage... x86 is older than the NES, about as old as the Atari, and it is the only component on a current PC that hasn't seen a true clean sheet modern redesign.

But why dump a mature platform for a newer one that isn't close to the same performance? At least wait until ARM has caught up before setting everyone back decades.

 

[mituzora]

I mean PCs and consoles should transition to ARM, just like the Macs did, as well as the Nintendo Switch... I am also not a fan of mobile gaming.
x86 is an almost 50 years old Frankenstein and is a hinderance to mankind at this stage... x86 is older than the NES, about as old as the Atari, and it is the only component on a current PC that hasn't seen a true clean sheet modern redesign.

if age is your only argument here, then I hate to break it to you, but ARM has been around since 32-bit x86 was introduced (1985). ARM used to stand for Acorn Risc Machine and was being developed for the Acorn computers.

I think ARM is great for mobile platforms, but when it comes to game development, it's not all that great unless you build something for it from the ground up. x86 is being used for consoles now because of the ease of porting and development. and x86 is fine, it's just power hungry. that "50" year legacy is incredibly important because of how insanely mature it is.

With as old as ARM is, it took a really long time to take off. Sure it's been being used in portable stuff for the last 20 or so years, but it's only just now becoming viable as a Desktop alternative. Imagine how long it would take to port 30-50 years of workflow, APIs, etc. to an entirely different architecture.

With all of that said, I don't hate ARM, but I think x86 isn't going anywhere any time soon, and for what it does, it's pretty good. it just doesn't last as long as an ARM CPU does on battery. that's literally it's only major downside in our current state of affairs.

 

[wiindsurf]

@subcon959

Performance is fine, just look at the Macs and their M chips.

And once the market sets their sights on ARM for desktop (which AMD, Nvidia and Qualcomm are rumored to be doing), the flexibility of ARM chip design means we could see an interesting shift in paradigms, with real technical innovation becoming possible again (as opposed to just timely IPC gains).

Do you recall the excitement that new console architectures brought in? And how dull the market has become?


@mituzora

I suspected this would be brought up, but ARM v8 is a clean new design from 2011, that fully breaks backwards compatibility, and restarts fresh.

I appreciate x86's BC, it is actually an amazing feat in itself, but that has only been possible due a patched up evolution full of compromises and hacks. The instruction set is humongous and makes decoding complicated (think speed and security) and well as expensive (chip real estate), the ring security model is archaic, parallelism is hard (think security again), and to this day there are still bits of undocumented instruction set behaviour. It is also too proprietary from a market point of view, making competition from third parties next to impossible, and disincentivising real innovation. Just think how much evolutionary opportunity we missed in the last 15 years when Intel was just stalling the market to the ground, just because it could?

I'm also not saying that I think x86 will die anytime soon, but that in my opinion it is due time for it to quietly ride into the sunset, for humanity's sake, if I may say. Also not saying ARM is perfect, but it is the closest real marketable alternative we have for the forseeable future, and it's openess in design possibilities and in market competition would be a real plus to the whole industry.

Post automatically merged: Aug 17, 2024

Macs are the absolute worst platform for gaming for a reason, and Switch is severely lacking in processing power. ARM chips are great for lower power consumption, but they're still playing catch-up with x86 CPUs in every other aspect, otherwise that transition would've happened already. Die shrinkage is what will determine the limit for x86 performance, and we're still nowhere close to hitting that limit.

Macs haven't been associated with gaming even when they were x86. And the Switch is underpowered by design choice. The transition never happened due to market forces, nothing to do with technical prowess. And yes, we are very close to hitting that limit you speak of, and it is the very reason why Intel is struggling at the moment, with even a section of the industry now considering Moore's law dead, including Nvidia's Jensen Huang.

 

[mituzora]

I'm also not saying that I think x86 will die anytime soon, but that in my opinion it is due time for it to quietly ride into the sunset, for humanity's sake, if I may say. Also not saying ARM is perfect, but it is the closest real marketable alternative we have for the forseeable future, and it's openess in design possibilities and in market competition would be a real plus to the whole industry.

I agree with this. eventually x86 will end up being left behind, but it's not gonna be any time soon, especially in the gaming market. This is probably by far one of the market segments that legacy compatibility benefits from the most. I also think that high-end PCs will likely stick with x86 for a long time, as the primary benefit to ARM doesn't affect something so heavily that's hooked up to mains power 24/7. Thankfully there are strides in this segment, but it's still not ready for the prime-time yet. (look at Jeff Geerling's videos on the ARM monster PCs he's demonstrated)

Just think how much evolutionary opportunity we missed in the last 15 years when Intel was just stalling the market to the ground, just because it could?

I don't think this is necessarily Intel's fault here. RISC, MIPS, ARM, etc. have been on the market for awhile now. I blame Windows and it's sheer dominance and reliance on x86. That doesn't mean though that Intel hadn't had a hand in on the stagnation of the market, I just don't think it's entirely their fault. Intel even tried to make an entirely new architecture (Intel Itanium). it just didn't take off. with that logic, we can blame AMD for extending the lifecycle of x86 with AMD64 LMAO

Honestly, I'd rather see RISC-V take off over ARM or x86 due to it's open nature as ARM still requires licensing, and has it's own proprietary BS due to chip manufacturers (looking at you Apple). But I can only dream lol. I'm all for other architectures taking off and overtaking x86, I just don't think it's gonna happen anytime soon. just too much legacy crap relies on it. Windows in its own right is over-encumbered with legacy code. If they really want ARM to take off, they're gonna have to find a half-ass reliable and performant translation layer akin to something like Rosetta 2, but I bet Rosetta 2 leverages certain proprietary bits baked into their silicon that Windows and Qualcomm simply hasn't done.

EDIT: another thing in Apple's corner is the fact that they've never really cared to hold on to legacy software anyways. PowerPC's translation layer didn't last for very long because the nature of that platform, and I can definitely see Rosetta 2 taking the same course; Most major software on Apple's platform has already made the transition to Apple silicon.

 

[Xzi]

The transition never happened due to market forces, nothing to do with technical prowess.

If ARM chips rivaled the performance of x86 CPUs, with much lower power consumption, anybody and everybody building their own PCs would be using them. "Market forces" would have no say in the matter. The fact is that porting to ARM always requires significant compromises in both graphical fidelity and performance, even when the chip in question is running at the exact same clock speed as the x86 CPU requirement. Gonna be a long time until even most PS4-era games are ported to Switch 2, let alone anything newer. Not to mention that means asking people to buy old games at full price, even if they've already purchased them before. Tough sell.

 

[mituzora]

If ARM chips rivaled the performance of x86 CPUs, with much lower power consumption, anybody and everybody building their own PCs would be using them. "Market forces" would have no say in the matter. The fact is that porting to ARM always requires significant compromises in both graphical fidelity and performance, even when the chip in question is running at the exact same clock speed as the x86 CPU requirement. Gonna be a long time until even most PS4-era games are ported to Switch 2, let alone anything newer.

to be totally fair here, part of that graphical fidelity drop isn't ARM itself per se, but more the GPU built into the SoC. If someone built an ARM based motherboard with PCI-E slots, and made drivers available for GPUs, or a better onboard GPU (again, look at Jeff Geerling's videos on the ARM desktops he's tested out) that won't be as much of an issue. but it did take a long time for ARM-based CPUs to be viable in use for Desktop/Laptop class machines. Apple silicon may not perform the best in gaming, but other GPU oriented tasks, it can handle very well. If Windows on ARM takes off, then there may be more of an incentive to develop better GPUs and/or better drivers for existing hardware.

 

[wiindsurf]

@mituzora

Nice, I agree with you on AMD64 putting x86 on extended life support, always thought the irony in that is kinda hilarious hehe

I used to think like you in wishing for RISC-V, or POWER now that it is fully open, but maybe I'm going through a more pragmatic phase at the moment, specially with the momentum ARM is getting lately... But yeah, one can dream right?

Talking about dreaming, I am a big fan of the PS3 Cell architecture. I know it had its problems (mainly the ring bus design, and perhaps the tooling around it was also not really where it should have been), but I think its true beauty lied in trying to let the hardware figure out paralelism / threading and taking that responsibility off the programmer. But perhaps it was way ahead of its time, requiring a big shift in mindset and workflow and during a period where AAA gaming was flourishing and big budgets and tight deadlines were more important than wasting time trying to figuring out how to properly program the damn weird thing...
Anyhow... would love to see an updated design based around that concept, as it would be very beneficial for security in general.

I agree with the Rosetta style stuff. Perhaps CPUs could also come with an optional frontend x86 decoder (something you could power on/off via BIOS), now that chips are becoming modular? Although I kinda cringe at that thought a bit as well... but if we have to compromise...