Is the SteamBox DoA?
Well, I read recently that Dell basically went ahead with what was to be their SteamOS offering and turned it into a Windows 8.1 HTPC effectively because the release of certain pieces of the SteamOS offerings were delayed.
At the time I didn't think much of it. Seemed like a smart move by Dell to make sure they aren't sitting on aging hardware any longer than they need to.
But Dell is one of the few SteamOS partners who actually could have afforded to sit on aging hardware and not worry. Some of the other OEM's in the game are much smaller fish. And as the release gets pushed back, they run the risk of losing their shirts in the process. If these smaller fish survive, do you think that they will still want to backup Valve in the future?
I wouldn't.
Rumours behind the delays I'm reading are primarily two things (disclosure: I have not done my homework, so these are purely rumours); 1 was that the controller wasn't ready yet and the other was that the graphics drivers weren't ready yet.
I would be unsurprised to find that both are true. On the controller front, Dell is shipping their system with Xbox 360 controllers, and as I pointed out long, long ago. Linux driver support for GPU's simply isn't anywhere near where it is for Windows or even Mac. Gabe Newell in his (lack of) infinite wisdom wanted to try and promote Linux as the next major gaming platform. Which of course, will never happen without proper GPU drivers on Linux.
This is why I said they were dumb to release it as OSS rather than a single preconfigured hardware option like the PS4 or Xbox One. Convincing a GPU maker to support one chip on one system was always much more likely than getting widespread adoption.
Worse though, recent moves by virtually everyone in the graphics processing scene are actually widening the gap. Technologies like Mantle, Metal and DirectX12 which also change the way the GPU's are leveraged in the first place are not coming to Linux that I've heard. And this means, even if Valve manages to promote the living hell out of Linux and get the drivers there. They'll STILL be behind on the rendering API's and the situation will actually be worse than it is today. Remember, Microsoft is claiming that DX12 games will basically double in speed on the same hardware. That is a bigger gap than exists in many cases for the drivers.
And, since that gain is multiplicative, if the drivers situation isn't resolved, the gap gets even wider.
At this very point in time, this is all conjecture. But, if Linux doesn't get find an alternative to OpenGL that runs as fast as DX12 which developers are willing to support, then when DX12 hits the streets and games start leveraging it, Linux based gaming will be derailed yet again (though to be fair, I don't think they were ever on the tracks).
And that will mean that once again, the only way for SteamBox owners to play the latest and greatest games will be to purchase both a SteamBox and a gaming PC.
Call it hedging their bets by allowing that streaming functionality. But I call it their death blow. Newell likely thought it was a decent stop-gap to keep people happy while Linux caught up to Windows in the drivers department. But with their failed attempt to deliver on time, and the changing landscape, that bridge back to Windows just gives developers even less incentive to target Linux.
If Valve is smart, they will release a set-top box whose exclusive job is to stream from a gaming PC to the TV and costs under $100.
If they don't move on that quick enough, they may find Microsoft has found a way to embed something like that into the Xbox or make Windows a better HTPC option than SteamOS. In other words, that bridge back to Windows may just end up buying Microsoft the time to completely kill SteamOS.
A lightweight and cheap streaming box that doesn't pretend that Linux will be able to keep up with the advances in GPU's might be enough to stop people from taking any attempt Microsoft makes at either HTPC's or game streaming seriously.
Basically, Valve dropped the gauntlets before they were prepared for a fight. Everyone has seen their hand, and they can either choose now to safely ignore it, or play a better hand and wipe them off the face of the earth. No, Valve won't disappear. Steam is a cash cow and failing in this segment won't do anything substantial to the company as a whole. At least, not in the short run. I feel sorry for the devs Valve hired for this fruitless task. The OS is probably going to be quite good. But lack of both popular and hardware support are likely to end it prematurely.
At the time I didn't think much of it. Seemed like a smart move by Dell to make sure they aren't sitting on aging hardware any longer than they need to.
But Dell is one of the few SteamOS partners who actually could have afforded to sit on aging hardware and not worry. Some of the other OEM's in the game are much smaller fish. And as the release gets pushed back, they run the risk of losing their shirts in the process. If these smaller fish survive, do you think that they will still want to backup Valve in the future?
I wouldn't.
Rumours behind the delays I'm reading are primarily two things (disclosure: I have not done my homework, so these are purely rumours); 1 was that the controller wasn't ready yet and the other was that the graphics drivers weren't ready yet.
I would be unsurprised to find that both are true. On the controller front, Dell is shipping their system with Xbox 360 controllers, and as I pointed out long, long ago. Linux driver support for GPU's simply isn't anywhere near where it is for Windows or even Mac. Gabe Newell in his (lack of) infinite wisdom wanted to try and promote Linux as the next major gaming platform. Which of course, will never happen without proper GPU drivers on Linux.
This is why I said they were dumb to release it as OSS rather than a single preconfigured hardware option like the PS4 or Xbox One. Convincing a GPU maker to support one chip on one system was always much more likely than getting widespread adoption.
Worse though, recent moves by virtually everyone in the graphics processing scene are actually widening the gap. Technologies like Mantle, Metal and DirectX12 which also change the way the GPU's are leveraged in the first place are not coming to Linux that I've heard. And this means, even if Valve manages to promote the living hell out of Linux and get the drivers there. They'll STILL be behind on the rendering API's and the situation will actually be worse than it is today. Remember, Microsoft is claiming that DX12 games will basically double in speed on the same hardware. That is a bigger gap than exists in many cases for the drivers.
And, since that gain is multiplicative, if the drivers situation isn't resolved, the gap gets even wider.
At this very point in time, this is all conjecture. But, if Linux doesn't get find an alternative to OpenGL that runs as fast as DX12 which developers are willing to support, then when DX12 hits the streets and games start leveraging it, Linux based gaming will be derailed yet again (though to be fair, I don't think they were ever on the tracks).
And that will mean that once again, the only way for SteamBox owners to play the latest and greatest games will be to purchase both a SteamBox and a gaming PC.
Call it hedging their bets by allowing that streaming functionality. But I call it their death blow. Newell likely thought it was a decent stop-gap to keep people happy while Linux caught up to Windows in the drivers department. But with their failed attempt to deliver on time, and the changing landscape, that bridge back to Windows just gives developers even less incentive to target Linux.
If Valve is smart, they will release a set-top box whose exclusive job is to stream from a gaming PC to the TV and costs under $100.
If they don't move on that quick enough, they may find Microsoft has found a way to embed something like that into the Xbox or make Windows a better HTPC option than SteamOS. In other words, that bridge back to Windows may just end up buying Microsoft the time to completely kill SteamOS.
A lightweight and cheap streaming box that doesn't pretend that Linux will be able to keep up with the advances in GPU's might be enough to stop people from taking any attempt Microsoft makes at either HTPC's or game streaming seriously.
Basically, Valve dropped the gauntlets before they were prepared for a fight. Everyone has seen their hand, and they can either choose now to safely ignore it, or play a better hand and wipe them off the face of the earth. No, Valve won't disappear. Steam is a cash cow and failing in this segment won't do anything substantial to the company as a whole. At least, not in the short run. I feel sorry for the devs Valve hired for this fruitless task. The OS is probably going to be quite good. But lack of both popular and hardware support are likely to end it prematurely.
Comments
Post a Comment