To make steamboxes cheap and with good enough quality graphics to compete with the PS4 or a high end PC they're going to need to sell the boxes at cost (or maybe subsidize) and cut great deals with hardware manufacturers.
I don't know if this can square with their plan of having third parties sell Steam boxes unless they are going to give these companies a cut of game sales or something like that.
Actually, a $4-500 PC can outperform/match these "next gen" consoles today. With parts bought from a store, so assuming Valve can get the parts cheaper, they can assemble a better performing machine for less than $400.
If you're wondering when PC hardware can "catch up" to console hardware, history shows it's simply a matter of waiting a year or two. Rather than the traditional new console generation land grab, Valve is playing a long game; they can afford to let Moore's law catch up a bit.
You can build, for the price of a ps4, a more powerful x86 computer than a ps4, today.
It isn't about two years, it is now. This console generation isn't selling at a loss, so all that is really happening is that MS and Sony have developed two entirely proprietary OSes with proprietary APIs and locked down distribution and usage, when under the hood they are just normal x86 based systems.
I think it really shows the absurdity of modern consoles in general. The industry made sense from the 70s to 90s when you needed heavily specialized hardware with finely tuned pipelines and exact design specs to be able to do real time video in 2 and eventually 3 dimensions, but today any computer can do that for fairly cheap, and console technologies are not entirely novel hardware concepts, but just cpus and gpus bundled on a board just like every other modern computer system.
If you want an argument for SteamOS, it is that coupling physical parts with a proprietary gaming platform is unnecessary and ridiculous today. So SteamOS doesn't do that. It is a console OS for any computer, in the same way the PS4 and Xbone operate, except you can just build your own rather than being shackled to hardware.
So you are still shackled to a proprietary software platform (Steam) but can console said platform anywhere it supports without being bound to silicon and copper.
Although not a gaming platform (a detail not to be dismissed), isn't coupling physical parts with a proprietary software platform Apple's recipe for success - and the direction Microsoft is headed, beyond the XBox business unit?
Microsoft initially took the decoupled route with desktops (sell software, primarily to OEMs), which obviously brought them great success; but the tides in this market are clearly changing. Could we be on the brink of a transition within the console market as well, albeit in the opposite direction (from closely coupled to decoupled).
If that is the case and given the fact that the mobile market is essentially closely coupled as well; any bets on a similar shift in 20 years there?
> isn't coupling physical parts with a proprietary software platform Apple's recipe for success
It is also Google's recipe for success with Android, since their Nexus products are always stellar. But by having Android the OS be independent of Nexus the Google hardware, they also have a magnitude more device sales per quarter running their platform than Apples.
And trying to directly compare Apple and Google revenue isn't really accurate since Google isn't trying to use Nexus or Android as a revenue generator.
Microsoft I feel is heading in the Apple direction out a lack of direction and lack of confidence in their own ability to innovate at all. But I don't think that is necessarily saying a wholly proprietary vertical stack in house is the absolute truth - it is like how the dark side is the easy way, but the light side wins in the end. It is easy for an industry to walk the single stack single company monopoly platform because then you can dodge the support across devices / OSes / etc angle, but in the long run having multiple devices / OSes / software stacks means competition, full market coverage, organic growth, and much more innovation, at least to me.
That doesn't necessarily change the economics so much. As time goes on a console manufacturer can drop the price of it's console and make up the difference using it's massive economies of scale and by selling more games (since it gets a cut of every title sold).
OTOH a hypothetical steambox manufacturer must still make a profit on every box sold. Since roughly the same pile of hardware can be sold as either a "steambox" or a "PC" (which can have steam installed) the margins on the two will have to be roughly the same.
Bare in mind that margins on most PCs are already pretty slim.
Bear in mind this is additional demand. People were never going to buy the equivalent "cheap pc gaming rig" because it didn't look and act like a console. Whereas steambox will, so effectively existing PC manufacturers will suddenly get access to a potentially huge new market which was off limits to them previously.
Given that the new console generation lifecycle will be about a decade, in 3-5 years SteamBox will be a no brainer in terms of price/performance.
I actually think what comes next is a "battle of the stores", where Steam will come installed on every Steambox but manufacturers will probably get paid to install other stores as well.
However, a "steambox builder" has more leverage for changing parts suppliers in order to cut costs; a console is absolutely stuck with the choices it made at the outset.
I don't know if this can square with their plan of having third parties sell Steam boxes unless they are going to give these companies a cut of game sales or something like that.