Valve's future console plans

[tuna]

I think it's easy to forget that the hardware side of PC gaming is an increasingly expensive hobby. As a problem Valve's customers (and potential customers) have, it seems to me like a worthwhile problem to solve with a Steambox.

Or MS to solve by allowing a Windows option on the Series consoles.

And let people buy games from Steam!

 

[PSman1700]

Valve entering the traditional console market? Probably not. What i could see them doing is make a stationary version of the deck with more power/capabilities. Just another option for the pc gaming market (which could sway console gamers to the platfrom) but not a tradtional console like the PS5/Xbox (and everything around it).

 

[Deleted member 11852]

I am the only person who remembers Valve's Steam Machines?

I don't see Valve going back to that failed experiment.

 

[cheapchips]

I am the only person who remembers Valve's Steam Machines?

I don't see Valve going back to that failed experiment.

Never forget!

Post Steamdeck & Proton we're in a very different world to the Steam Machine experiment. Valve doing their own hardware would mean it shouldn't suck, either from a price or design perspective. All the Steam Machines failed on one or both of these fronts.

Proton is doing wonders in ensuring the Stream library is mostly available, as opposed to the limited number of Linux titles Steam Machines offered. SteamOS is also offering more 'PC gaming done right' with it's system wide settings, fps limiters, shader precompile etc.

 

[SpaceBeer]

Console warring is another thing all together. And you never know how MS may respond to such competition. Consoles is a competitive space with a number of players so it would hard to argue that an approach by MS to actively and aggressively break compatibility on the DX side with Proton as anti-competitive.

Steam deck might solve "chicken or egg" problem for linux gaming. So in future it might be the case to see more native linux games, as developers/studios would expect it to work better than proton/wine. If Deck sells in millions (and it might be the case), you know that 5-10% of your users are on Linux, then why wouldn't you invest in the platform.

If I was Phil Spencer I would really think about how I would avoid being Nokiad. Microsoft is kind of boxed in inside their platform while Valve has a unique opportunity to expand theirs.

There is no reason for vast majority of gamers (PC users) to move away from Windows, so as long as PC gaming lives, Microsoft (Windows, DirectX...) will be the leader. And whatever "PC console" comes to the market, it will not compete against Xbox nor standard gaming PC.

IMO, if M$ tries to break D3D in order to prevent gaming on Deck (Linux) it would be lose-lose situation for them

 

[Silent_Buddha]

IMO, if M$ tries to break D3D in order to prevent gaming on Deck (Linux) it would be lose-lose situation for them

Fortunately, there's no reason that MS would desire to do this with the current management who have mostly embraced Linux on Windows. As well, there's little to no financial reason for them to prevent games running on Linux as the Microsoft Store is a minor player WRT game sales. Game Pass which will likely continue to require Windows on PC would be viewed as more than enough to ensure that Windows remains the primary PC gaming platform. Hell, just Linux's general user unfriendliness if you aren't very familiar with hardware, command line interfaces, and being able and willing to potentially spend hours looking up solutions to problems in Linux online is enough to likely ensure that Windows remains the primary gaming platform on PC.

All that said, I could even see a potential future where MS could release a Linux Game Pass client.

Regards,
SB

 

[eastmen]

Here are my thoughts on this.

I think they should continue with steam deck and optimize around it. They should launch a steam deck 2 , perhaps when amd moves to a mobile 15w zen + rdna 3 .

However I think the "home console" should simply just be the steam decks without a screen and battery with a revamped steam deck layout controller.

While I think they could push out steam consoles using whatever amd set up or even intel/ nvidia set up they will eventually run into the same issues traditional pc gaming has. There will be users with so many variations of these devices that steam deck or steam console verfiied will loose all meaning. By keeping with a smaller amount of hardware combos it will allow devs to create optimized modes.

So you can have steam deck optimzed performance mode , steam deck optimized fidelity and so on. if they launched a steam deck console (no screen/ battery) you could then allow steam deck console optimizations or mostly fidelity upscaling to hit 1080p or whatever.

Introduce steam deck 2 and a user can use steam deck optimized settings most likely moving over to fidelity modes and devs can do another pass to make steam deck 2 optimized settings and newer games could have steam deck 2 optimized. Again add a home version of a steam deck with no screen or battery and add upscaled versions for those.

I'd guess there would be 3-4 years between models and so developers don't become over whelmed with optimization targets

 

[cheapchips]

However I think the "home console" should simply just be the steam decks without a screen and battery with a revamped steam deck layout controller.

I love how compact such a device would be. You'd be able to cram it in a Steamlink* sized box. Really underpowered for even 1080p gaming though. I'd prefer that they had at least parity with current gen consoles.

* For those that don't own one, despite only being the size of a couple of playing card boxes, Steamlink weighed an absolute ton. I think the case is carved out of a neutron star.

 

[Deleted member 11852]

Post Steamdeck & Proton we're in a very different world to the Steam Machine experiment. Valve doing their own hardware would mean it shouldn't suck, either from a price or design perspective. All the Steam Machines failed on one or both of these fronts.

I think the biggest barrier is that it will be difficult for Valve to be competitive on the price of the hardware when they are competing with bigger players in a commodity market. As others have said, if Valve are not looking to make money on the hardware, because they already make the cream of profits with Steam, why even bother with designing, manufacturing, shipping, selling and supporting hardware?

SteamDeck filled a niche, a gap in the market.

 

[cheapchips]

I think the biggest barrier is that it will be difficult for Valve to be competitive on the price of the hardware when they are competing with bigger players in a commodity market. As others have said, if Valve are not looking to make money on the hardware, because they already make the cream of profits with Steam, why even bother with designing, manufacturing, shipping, selling and supporting hardware?

SteamDeck filled a niche, a gap in the market.

On the whole, I wouldn't see it as competing with consoles. It's about continuing what the Deck delivers, a more accessible way into PC gaming (well, Steam gaming).

Work and educational laptops are rubbish gaming devices. They can't be nudge upgraded for gaming in the way desktops of old used to be. The price of dedicated gaming PCs are a huge barrier to entry. You have to be very serious about tiles also available on cheaper platforms or just like building stuff.

You can't buy a reasonably priced gaming PC. Doesn't matter if it's going in the office, bedroom or livingroom. The Deck is as good as it gets and it's going to hit a wall with current gen titles. It's a perverse situation when XSX/PS5 are essentially PCs, with locked down software and storage that was slightly forwards looking at inception.

Obviously Valve have all the data as to whether it's really worth a punt for retaining/acquiring customers.

 

[Deleted member 11852]

@cheapchips so what are the defining qualities here? cheap, upgradable and competitive (performance?) with consoles?. I think this is one of those product triangles where you only get pick two things! I.e. you have high quality work, fast work and cheap work. You're never going to get fast, cheap, high quality work ;-)

And if you're going Steam only, you're giving up Fortnite outside of streaming.

 

[eastmen]

I love how compact such a device would be. You'd be able to cram it in a Steamlink* sized box. Really underpowered for even 1080p gaming though. I'd prefer that they had at least parity with current gen consoles.

* For those that don't own one, despite only being the size of a couple of playing card boxes, Steamlink weighed an absolute ton. I think the case is carved out of a neutron star.

Like I said the issue would be that you'd keep adding more configurations to optimize for which will drastically reduce the likely hood of any of them being optimized for.

 

[cheapchips]

Like I said the issue would be that you'd keep adding more configurations to optimize for which will drastically reduce the likely hood of any of them being optimized for.

Is optimisation that much of an issue, so long as it works on Proton/SteamOS?

 

[eastmen]

Is optimisation that much of an issue, so long as it works on Proton/SteamOS?

For the wider public I would imagine it would be.

You buy a switch and the games just work
You buy a xbox series or a ps5 and you choose if you want performance or fidelity.

You load up the steam console and suddenly your futzing around with settings trying to get games to run ? At that point you just have a pc and valve already tried that and failed.

 

[Deleted member 11852]

Is optimisation that much of an issue, so long as it works on Proton/SteamOS?

If by "accessible" you mean readily affordable then you're already looking at targeting a modest hardware specification and any software that doesn't come with some level of TLC to get it working well is inevitably going to struggle.

This is the only reason that consoles have been been a viable alternative to the PC for generations. The idea that you could run Crysis on a 360/PS3 with under 512mb RAM actually accessible is because some poor buggers slaved for weeks/months to make it work on that configuration. Or GTA III on a 32mb PS2 when the PC version required 128mb + video RAM.

 

[cheapchips]

If by "accessible" you mean readily affordable then you're already looking at targeting a modest hardware specification and any software that doesn't come with some level of TLC to get it working well is inevitably going to struggle.

This is the only reason that consoles have been been a viable alternative to the PC for generations. The idea that you could run Crysis on a 360/PS3 with under 512mb RAM actually accessible is because some poor buggers slaved for weeks/months to make it work on that configuration. Or GTA III on a 32mb PS2 when the PC version required 128mb + video RAM.

I'm not convinced that performance paradigm holds true anymore. AAA games aren't made for PC and squeezed into consoles. They're made for consoles and then PC gamers turn on RTX + ultra settings 'cos they can. The non AAA PC space tends not to require flashy hardware.

Current gen consoles are essentially PC hardware, and aging hardware at that. Look as well at what Steamdeck wrings out of it's dinky APU and compatability layer.

If Valve is willing to commission a 350mm² 4-5nm APU, I'd bet on it returning pretty good performance on console+ settings.

 

[Deleted member 11852]

I'm not convinced that performance paradigm holds true anymore. AAA games aren't made for PC and squeezed into consoles. They're made for consoles and then PC gamers turn on RTX + ultra settings 'cos they can. The non AAA PC space tends not to require flashy hardware.

The not-quite consistently perfect performance of PS5 and Series S and X against various PC configurations I think does demonstrate how much more performance is possible from a very cheap console compared to a PC costing even just 50% more on the hardware. And that's with most games running on DirectX natively on Windows with no API translation layer that SteamOS relies on.

Current gen consoles are essentially PC hardware, and aging hardware at that. Look as well at what Steamdeck wrings out of it's dinky APU and compatability layer.

SteamDeck struggles to run Marvel's Spider-Man at PS4 performance levels. PS4 was weak hardware configuration even in 2013, almost a decade ago.

If Valve is willing to commission a 350mm² 4-5nm APU, I'd bet on it returning pretty good performance on console+ settings.

Expensive work. How are they going to recoup the cost? Sony and Microsoft do this by selling tens of million bits of hardware then leveraging licensing. Steam has no licensing for SteamDeck, they rely only on the revenue cut of Steam - which Sony and Microsoft also have in addition to licensing and masses sales of juicy profitably accessories like controllers, headsets and other bobbins.

The video console economy is viable due to being able to provide very cheap hardware, recouping costs from licensing and sales on tens of million of hardware units, and accessories. That's how all the R&D and hardware manufacturing is made viable.

 

[eastmen]

The not-quite consistently perfect performance of PS5 and Series S and X against various PC configurations I think does demonstrate how much more performance is possible from a very cheap console compared to a PC costing even just 50% more on the hardware. And that's with most games running on DirectX natively on Windows with no API translation layer that SteamOS relies on.



SteamDeck struggles to run Marvel's Spider-Man at PS4 performance levels. PS4 was weak hardware configuration even in 2013, almost a decade ago.


Expensive work. How are they going to recoup the cost? Sony and Microsoft do this by selling tens of million bits of hardware then leveraging licensing. Steam has no licensing for SteamDeck, they rely only on the revenue cut of Steam - which Sony and Microsoft also have in addition to licensing and masses sales of juicy profitably accessories like controllers, headsets and other bobbins.

The video console economy is viable due to being able to provide very cheap hardware, recouping costs from licensing and sales on tens of million of hardware units, and accessories. That's how all the R&D and hardware manufacturing is made viable.

Um what ?

Marvel's Spider-Man Remastered on Steam Deck: Performance, Settings, AyaNeo Air Comparison!

Marvel's Spider-Man is perhaps one or two patches away from PC perfection, but already the game has super scalability f…

www.digitalfoundry.net

The steam deck runs between the ps4 and ps5 even with the engine having some issues with amd cpus.

 

[Deleted member 11852]

The steam deck runs between the ps4 and ps5 even with the engine having some issues with amd cpus.

I saw the earlier analysis and it looked like the SteamDeck was struggling to maintain 30fps PS4 performance targets, i.e. 1080p.

Steam Deck? Using tweaks to the high preset - crucially, medium hair quality and low depth of field - you can essentially run the game at 720p/800p at a fairly consistent 30 frames per second, using dynamic resolution scaling to help mitigate areas where the hardware is GPU-limited.

 

[cheapchips]

Expensive work. How are they going to recoup the cost?

A custom APU was worthwhile for whatever binned surface device Steamdeck's APU was originally destined for. It is a smallish chip though.

Tesla's an interesting volume reference point, if not a cost one. Their 10 TFLOP RDNA2 APU is only destined for a few million cars a year.

AMD should be really well geared for Zen cores + your choice of CUs by now.