Certification could give Steam Machines more appeal through easily grasped internals.
by Kyle Orland
by Kyle Orland
Talking over Valve's announcement of its Steam Machine prototype specs with a few people online (including Ars' own Andrew Cunningham), I've come to the conclusion that Valve might need more than its own free, standardized gaming OS (and, ideally, an exclusive killer app) to make PC gaming appealing for the living room console consumer. To really put up a fight, they should do something to simplify the dizzying variety of architectures and performance points that are inherent in parcel with PC gaming.
Just look at the range of hardware configurations that Valve is including in the roll out of its Steam Machine prototypes. They range from machines with parts totaling $600 or so to Nvidia Titan-powered beasts that would cost upwards of $1500 to build. Valve is also quick to mention that plenty of other companies will be rolling their own systems to market, some of which may deviate "substantially" from the prototype. Yet all of these different configurations are falling under a single "Steam Machines" label.
PC gamers are used to the fact that the term "PC" covers everything from $200 budget machines to $3,500 beasts. Valve is explicitly targeting the living room-based, console-playing audience with its SteamOS boxes though. If this audience is used to one thing, it's having a single standard that provides consistent, measurable hardware power levels and compatibility.
There are plenty of advantages and disadvantages to both sides of the great console/PC gaming divide, but this consistency is one of the console side's enduring pluses. You can be sure that any Xbox 360 will run any game that has come out for the system between 2005 and this holiday season, regardless of slight differences in the internal hardware. With the PC side on the other hand, you have to worry about everything from what APIs your graphics card supports and how much RAM the game requires to how many cores and per-second cycles your CPU can handle. Not only that, but you have to juggle minimum, recommended, and optimum system requirements for every single game.
The advantage of this setup, of course, is flexibility. You know exactly what piece (or pieces) of hardware you'll need to replace to get your system up to snuff for the latest game. But the console gamers Valve is newly targeting, by and large, just want to know that anything made for that particular system will run. They don't want to possess deep knowledge of their system's innards and API calls.
Valve doesn't have to compromise on SteamOS' open, freely upgradeable promise to reach out to these gamers. It just has to offer up some sort of standardized system for separating marketable Steam Machines from a variety of vendors into different power tiers that developers and consumers can easily target.
Under this system, Valve would set different levels of minimum performance that a SteamOS-machine would have to meet to earn a Valve-endorsed certification at Level 1, Level 2, and so on. Game developers, in turn, could optimize and advertise their SteamOS-powered games with a specific level in mind. Maybe the game runs best on a level-4-or-better system but can still run acceptably on any level 2 system.
This setup would lose some of the granularity associated with the current "system requirements" regime common in PC gaming (a game that had the RAM requirements of level 5 but the GPU requirements of level 2, for instance, would be forced to simply "require" level 5 under this setup). But it has the advantage of being much simpler than the lists of system requirements and the wide variety of components that PC gamers have to deal with today.
If a game is tuned for Level 5 Steam Machines and you have a level 6 Steam Machine, that's really all you need to know. No need to squint at the side of a box or go hunting around the Web for details. If your machine is below the minimum level of most of the new games you're interested in, then it's time to upgrade. For an audience used to simple, plug-and-play, "it just works" compatibility across an entire hardware generation, this kind of simplicity could be key to adoption.
Valve would ideally set the standard for each of these tiers by releasing its own Steam machines at various levels (its prototype plan already shows a willingness to support many different hardware configurations). This would give outside Steam Machine makers a good reference to shoot for in developing their own hardware, much as Google's Nexus line helps set the standard for the Android market. Having discrete tiers would also help consumers comparison shopping for a Steam Machine determine how much power they're getting for their comparative dollar.
System makers could opt out of this certification of course, but if Valve starts including it on its own hardware and promoting it on Steam itself, consumers may well start looking for that Valve-endorsed certification sticker on other Steam Machines. Do-it-yourself Steam Machine builders could use a website to determine their hardware's current certification level and to measure the potential impact of upgrading one piece or another.
Microsoft has tried something similar with its Windows Experience Index scores, which were meant to give people some idea of which components in their PCs were fastest and slowest. Those numbers never really ended up catching on, but Valve is in a unique position to use its launch of a brand new gaming platform to roll out a standard that could be widely adopted across the industry.
There's a lot we don't know about Valve's SteamOS plans, and they might have something like this in the works already. Regardless, this kind of centralized standardization would go a long way towards getting console gamers used to the idea of a PC-based system—without giving up any of the flexibility and upgradeability inherent in the SteamOS design.
0 comments:
Post a Comment