Steam Machines: Avoiding 3DO’s Destiny

We expect the Steam Machines to come on the market in 2015. For many Linux supporters, this sounds like “the year of the Linux (gaming) desktop” all over again, however there are a number of hurdles to adoption that Valve and their partners will have to tackle. And seriously, it is not going to be easy, no matter how you look at it.

First, the idea of a console using a standard and a common OS, manufactured by different companies, is not really something THAT new. That’s actually very much what Trip Hawkins, founder of EA and subsequent founder of 3DO, tried to achieve with the 3DO standard. 3DO acted as a independent company responsible for designing the OS and hardware specifications, leaving the manufacturing, pricing and marketing to other companies licensing its technology. The situation is a little different nowadays, but overall there are many learnings from the 3DO failure to win on the market that remain very relevant.


3DO’s first support came from Panasonic, which was looking to make a dent in the lucrative consoles market as the 32 bits market was expected to take over the aging 16 bits consoles in a matter of years. However, the 3DO was not so much positioned as a console in the first place. It was the age of everything multimedia, and the Panasonic model was released as a kind of VCR that could also do gaming… and it was priced accordingly… at like 700 USD at the time – which is more than 1000 USD in current dollars. This did not help to secure a strong position at launch, even though the Panasonic 3DO was early on the market before other 32 bits consoles. Positioning is probably not going to be main issue with Steam Machines, as we expect them to be marketed more or less as consoles (wait, Alienware is saying their Alpha can also be used as a Windows PC… “More than A Gaming Console: Alpha takes you where other consoles can’t; plug in a keyboard and mouse and transform your Alpha into a fully functional Windows 8 PC”), but pricing may be a real concern if Valve expects some penetration to larger, non-PC gaming audiences.


Pricing is a key element in the value proposition. Hardware wise, it’s likely that most Steam Machines out there supporting AAA games will need serious hardware, and therefore be reasonably pricey (because of the limited production volumes). On top of that, just like for the 3DO model (or the Android model for that matter), manufacturers do not make money on the Steam platform itself (that’s Valve’s territory). Their business model is all about doing margins on the hardware and their potential specific services. Modern console manufacturers like Sony or Microsoft, on the other hand, sell consoles at a reasonable price and recoup the costs on years of software sales where they secure licensing fees to generate profits. For the end user, this means that current-gen consoles will most probably stay the cheaper ways to play modern games, while Steam Machines (at least the ones with decent hardware) will remain more premium options. Back in the 90s, 3DO consoles remained quite expensive, despite Trip Hawkins’ will to have them decrease their price point. 3DO was unable to convince manufacturers to do so, and instead decided to finance more advertising budget by increasing the licenses costs for games developed on 3DO. Ouch. That did not work too well with third party developers. Valve is not in the same situation here, since their customers are already on PC and Mac and they do not need an extra hardware platform to survive. However if they expect some added business and growth from the Steam Machines, fair pricing for the end user will matter anyway.

Brand dilution

This was a major problem with 3DO. Who represents the brand? Which manufacturer stands for it? When you have 10 different Steam Machines out there, what is the meaning of a Steam Machine ? What is it supposed to be like ? Unlike phones which pretty much all look the same, some of the Steam Machines presented in early 2014 had virtually no common look and adopted radically different designs. This is very much like the PC Desktop Market, but very different from what people expect from a console standard. The 3DO had the same issue, the lack of consistency in the design and especially brand recognition. Valve is at least bringing SOME consistency here, by building a controller from scratch and letting manufacturers use it. That may be only common component between all options out there, while everyone else can recognize what a Xbox, WiiU or PS4 should look like.

Communication, Advertising

There is a key issue with the “everyone can make it” model: even if all manufacturers benefit from the shared platform, they have to nevertheless sell their own and the marketing costs are not shared at all, and this will lead to a fragmentation of advertising and communication. Sometimes this is detrimental, especially if major players are unable to effectively communicate the benefits or advantages of Steam Machines versus other systems. And since most companies set to produce Steam Machines are in the business of assembling hardware, their forte is certainly not marketing. I wonder if Valve plans to do something around that issue, and if they intend to lead the way to avoid replication of efforts and ensure some consistency. I don’t think they will, as they clearly want to distance themselves from that role.


Classification of Hardware

While not related to the 3DO case per se, another major issue that Steam Machines will face on the market is the lack of user-friendly way to tell what games each of them can play. On regular consoles, there is a single hardware base and games run in the same way everywhere. A Steam Machine could however run on Intel HD graphics or nVidia GTX970 hardware. Or anything in the middle. How to tell users that to play Metro2033 your Intel HD base will not be sufficient, even though the game is available for the platform ? Current PC Gamers will have no issue to “get” that aspect right, but thinking beyond obvious users, your regular “know nothing about PCs” person will have no clue as to what they should get. So, either manufacturers will implement their own guidelines to help consumers (if they do it’s very likely to suck balls), or else Valve could step up and provide some kind of benchmarking solution to ensure gamers are clearly aware of what they purchase and what it can play. There’s many ways such a benchmarking system could work: assign target performance for a number of games and if the system meets them, it gets a grade A. If not, it ends up anywhere between grade B to F. Add the year number to that benchmark, and you end up with A2014, C2015 for example. A2014 being able to run all games up to 2014 perfectly in full details, while C2015 struggles with high end titles from 2015 while still remaining a good option for 2014 or older titles. When buying a game, Steam would then let someone know if their configuration is good enough to play it or not, to manage their expectations. That’s the kind of things SteamOS needs way more than tweaking its UI at the moment: how to handle the differences in hardware in the future while keeping it simple for everyone?

The Attack of the Clones

This is actually unrelated to the 3DO story, but this is going to be a major point to address as well. Steam Machines with SteamOS versus non-official Steam Machines running Steam on Windows. It is already happening, and faster than we thought. Alienware’s Alpha is about to be released, and will use Windows 8.1 as a back-end. As a consumer, once you have the two offers on the market, which one do you go for ? I can see the appeal of a Windows-based Steam console: you get access to a very large catalog of games, and all games will come first for that system. A SteamOS Machine (running Linux) will have a reduced library of titles to choose from. Why go for the later one ? Valve’s Steam Controller may be produced by other manufacturers as well, since the design schematics will be open, as announced. So, what is the real benefit of the SteamOS for the end user? Since Valve has on several occasions refuted the possibility of exclusive games on SteamOS, it’s far from being clear what will be the actual differentiators. And good luck explaining to Windows gamers who know Steam that the Steam Machine only plays about 10-20% of their games catalog. And for Valve, such machines are helping their business model anyway, so there’s no clear incentive for them to try to stop the Windows clones.


Valve has been fairly silent about the Steam Machines in the recent past. Especially about how they perceive the Alienware Alpha and upcoming clones. There were a couple of SteamOS updates but certainly nothing really significant. It could well be that Valve is behind schedule and realized that there is a lot of work down the road to reach a user-friendly solution. Or they are simply not feeling the urgency anymore. I have recently pointed out on the GNU/Linux Gaming subreddit that they will not unveil anything related to the Steam Machines for the CES 2015 (in January) while they plan to share more updates during the GDC. Skipping the CES means they are probably not ready to show anything to the public at this stage. Interpret this as you wish 🙂

BoilingSteam lets you access our content for free, but writing articles is a constant investment. We don't use ads or sponsporship, help us make our activities sustainable by donating via Patreon or LiberaPay if you prefer it anonymous. You can follow what we do via our newsletter, our RSS feed, our Mastodon profile or our Twitter feed. We also have Peertube, Youtube and LBRY channels. If you'd like to chat, you can also find us on (what is Matrix?)

5 1 vote
Article Rating
Notify of

Oldest Most Voted
Inline Feedbacks
View all comments

Valve is in it for the long term. They don’t need the steamboxes to sell x million in the first month. Sure, you need windows for gaming now, but who knows what state gaming/windows will be in 5 years from now? Look at the dent apple put in their marketshare. Linux has the potential to do that and more. Also, right now theres over 800 games available for linux through steam. What console launched with a library that big? All they really need is a big launch title. Imagine a Steambox + controller + Portal 2: episode 1 bundle, for… Read more »


> And good luck explaining to Windows gamers who know Steam that the Steam Machine only plays about 10-20% of their games catalog.

I don’t think this is a problem because I doubt PC gamers are the target audience of the Steam Machine. Valve is in the business of selling games, and PC gamers are already buying. The Steambox is for people who don’t have a gaming pc, for console gamers.

I agree with all your other points though, Valve has quite a bit of work to do to pull this off.

David Hollinger

While I don’t see Valve doing this, they could do what Google does with Android now and require a specific set of standards to be adopted to gain access to Valve-specific features. * Lose Steam Branding if the Machine doesn’t run SteamOS * Not allowed to pre-load Steam onto Windows PCs * Allowed to pre-load Steam, but not allowed to pre-configure the Windows PC to boot directly into Big Picture mode. * Not allowing the Controller to be SOLD with non-SteamOS Machines (Sure the schematics for the old one are online, but that doesn’t give rights to companies to profit/sell… Read more »


>Lose Steam Branding if the Machine doesn’t run SteamOS

I think they are doing this. Look at for example the Alienware Alpha it doesn’t have the steam logo it also is not marketed as a Steam Machine.