Steam Machines: A Very Quiet Launch


And so November 10 is now past. It was the official release date for SteamOS and the Steam Machines initiative. This was not like a regular console launch, with queues in the streets before the first sale goes off. This was, if anything, completely anticlimatic. Even the Steam Store did not bother so much about it. There was certainly a bunch of SteamOS compatible titles going on sale, but as far as sales go, there was nothing exceptional there.

And while I think it’s a very good thing to have SteamOS and Steam Machines out there, it feels a lot like “too little, too late” at this stage. For several reasons.

  • Most of the Steam Machines are not even available right now. Sure, you can get the Alienware Steam Machine, or maybe the ZOTAC Nen, but a lot of the other previously announced manufacturing partners actually delayed their hardware launch, or even cancelled it altogether : Origin has just announced they have decided against SteamOS for their upcoming Steam Machine, preferring Windows instead.
  • Most of the big AAA titles ports are very, very late. I’m not even talking about newly announced titles, rather the ones announced at the GDC back in March. We did get Shadow of Mordor (with extremely poor performance vs Windows), Alien Isolation, Companies of Heroes 2, but most of the other titles are missing: Grid Autosport, Payday2, Saints Row 4, Rome Total War 2, Witcher 3, Batman Arkham Knight (this one was botched on Windows so the delay is expected)… They had more than 6 months to get them out, and they failed. Not sure who is to blame, but it’s an extremely poor showing when stuff that was announced previously is still missing, with no actual release date in sight.
  • Worse, there was no new announcements following the Steam Machines launch. No new partners. No new AAA ports unveiled. No new games available released on that particular day. It’s like whoever is in charge of supporting the platform forgot what Marketing is about or something. There should always be some important announcements once in a while to show the platform is going in a positive direction, to keep interest up. November 10 was the right time to do so, yet nothing happened, once again.

The press did not take too long to find out what everyone in the Linux community already knew: most AAA titles performance way poorer on SteamOS vs their Windows equivalent, and Arse picked that up in a recent article. Unfortunately they picked Shadow of Modor for their banchmarks, which is one of the worst performing ports out there - if they had picked Dirt Showdown or Borderlands 2 the picture would have been a little brighter - yet it is a fact that SteamOS games tend to perform poorer, and while SteamOS is a tighly integrated environment for games, the following question is legitimate: “Why would I buy a Steam Machine that has less games and performs worse than a Windows box ?”. After Valve demonstrated that Left4Dead2 could run faster on Linux vs Windows, one of the potential benefits of SteamOS could have been higher framerates, but that train is long gone and at best we can expect parity nowadays, and worst comes to worst, 50% performance. It does not matter for old games, but new games requiring higher end hardware, it can clearly be a deal breaker for anyone who is not aware of what to expect.

To further increase the value proposition of Steam Machines, I hope that Valve and their friends continue to actively work on…

  • More AAA ports and same day releases as much as possible, from existing partners.
  • Securing new ports from other major publishers. Obvious ones include Bethesda and Rockstar, while they may be very hard to convince.
  • Bridging the performance gap on newer releases, either via better OpenGL practices, or the introduction of Vulkan as soon as 2016. It is however improbable that most AAA titles use Vulkan in 2016, so we still need OpenGL to deliver for a while no matter what.
  • Success stories. To build a great platform there needs to be a reason for it to exist. A problem it solves. Steam Machines bring new games on top of a Free Software OS, but that benefit does not matter for most end users. If games are not going to perform better, or be as many on the platform for the time being, maybe Valve should play on the openness of the platform and why there are advantages of having an open OS in the first place.

At this stage it’s still too early to say what will become of the Steam Machines. I would have a hard time to sell them right now, but in 1-2 years from now, provided they have way more games (games that people want to play, not 1000’s of crap that interests almost no one), and that the hardware gap starts to show between consoles and modern PC-based hardware, things should start to look a little more exciting. It’s clearly going to be a long game, let’s hope manufacturing partners do not lose patience and understand SteamOS needs time to grow.