The State of Steam Machines in June 2016

While this article may sound like an answer to the recent pieces seen on Ars, or PCWorld, regarding how poorly the Steam Machines have done in the past 6 months, this could not be further from the truth, since I had this article as draft since May and only decided to finalize it recently using the most recent data available following the disclosure of Valve and the 500 000 Steam Controllers sold. So, this one will try to answer the question “How are the Steam Machines doing, 7 months later?“.

There are several ways to look at it. Has there been any activity from the different manufacturers to release new machines? Any new entrants ? Can we estimate the sales so far ? How much cash did those manufacturers make so far ? Was it worth it for them ? We can try to answer all those questions, using some estimations.

First, in terms of activity, well, the data is kind of clear. 7 months later, there has been, to my knowledge, no major new entrants in the Steam Machines world. But there are small companies, like, which now offer their Steam Machine (basically typical Gaming PC in a small tower) for example at 799 EUR, even though Valve shows their machine as being not available, while it’s in stock on their side. What the Hell ?

Among the other small companies:

Hey, but there’s one additional option, the MainGear Drift, that I was not aware of. Pricey from 1099 dollars, but well… and let’s not forget the Scan 3XS ST machines, with decent options for GPUs.  I am not sure when those were made available, but apparently they were already sold back in November 2015.


The three “main ones” are still being sold and advertised on their manufacturer’s websites:

One of the problem I have is that… it seems like the launch page of the key Steam Machines (i.e. the Alienware one, that Valve even promoted in the first place) has never been updated since November 2015. Look at the info available right there:

Have a PC game that you want to play on your TV? Not a problem as you’ll be able to play over 1500 games locally or stream any of the 6000 Steam games on your home network directly to your TV.

1500 games? Guys, we passed 2000 games already…

MASSIVE GAMES LIBRARY – Choose from thousands of native SteamOS games and over 5,000 additional games available via streaming. Free copy of Rocket League and Portal 2 included in all pre-sale purchases.

Oh, maybe you should stop talking about Rocket League… as it’s a port in limbo for like 7 months already.

Yes. You’ll be able to stream TV shows and movies or listen to music while gaming. Steam delivers more than just games to your living room.

Oh, please do tell me know when that feature is coming in SteamOS, because it seems to be missing.

There are 3 Full version games: PAYDAY 2 2015 Game of the Year Edition (Coming to SteamOS in early 2016), Screencheat and Robot Roller-Derby Disco Dodgeball. Additionally users will get these items available in Brawlhalla and Warthunder, both free to play games: branded skin and 200 Mammoth Coins (Brawlhalla) and an M18 Black Cat Tank and XP-38G Lighting plane (Warthunder). There is also an Alienware exclusive demo of the game Coffin Dodgers.

Guys, we are not in “early 2016” anymore, it’s time to update your page.

Syber is not without blame since they proudly say the following about their Steam Machine…

Don’t limit yourself to the overpriced and limited selection of console games. With the SYBER PC Console, you’ll get access to any of the over 10,000 Microsoft Windows compatible game title from all major publishers. Access major gaming libraries including Steam, UPlay,, Origin and the Windows Games Library.

Again nothing new, that kind of problematic wording was already there at launch. Just, no change, no update.

This is, I think, telling that such companies are not seeing the sales they expected and do not care in investing anymore time in trying to market them further.

But what kind of sales are we talking about really ? As mentioned by Valve themselves, they sold about 500 000 Steam Controllers. Based on our own survey done within the GNU/Linux gaming community, we can roughly estimate that there’s a ratio of 1:25 between ownership of a Steam Machine and that of a Steam Controller. Of course, I have absolutely no idea if that ratio is valid across other, larger communities (it could be lower, or larger), but that’s a better estimation than none at all. How do we know, though, that this ratio is not completely utter bullshit ? Well, the Amazon reviews give us a hint: The Steam Controller gets about 1047 reviews at the time of writing, while the Alienware Steam Machines have about 38 reviews, which is a 1:27.5 ratio. Obviously we don’t know if the percentage of reviews is consistent across products, but in general it is more or less the case. So our ratio of 1:25 seems appropriate for this exercise.

For such a ratio that would mean that about 20 000 Steam Machines were sold. You can play around with the ratio… if the ratio is twice better, then you get 40 000 machines. Twice worse, and only 10 000 machines. Anyway, it seems to be in the tens of thousands, and this sounds like a reasonable guess.

As to who profited from these sales, I think I’m not mistaken by saying the large majority went to Alienware’s. There are several ways to tell that, we have our survey as one indicator (Alienware is the only one with high awareness – 95% have heard about it, while for Syber/Zotac the awareness drops to about 75%), we have the number of Amazon reviews as another. Again, just an estimation but I would say 6/10 bought an Alienware Steam Machine. So if we keep our prior estimation of 20 000 units,  12 000 units for Dell (Alienware’s mother company).

How much money do they make on such hardware? This is a much more difficult question to answer. One way to address that question is to look at the typical profit ratio of PC makers, and there are already pretty good estimations made out there before, since most PC makers are public companies and you can easily derive such metrics. It’s usually about 3%. But let’s assume they made some higher margins with this hardware, since they don’t make any money by installing crapware on such machines. Let’s feel generous and say they made 5% of profit. If we assume an average price of 500 USD for the different models of Alienware Steam Machines sold, that would be about 300 k USD. A third of a million dollars, give or take. Not bad, but still a drop in the ocean (even if we assume it was twice or three times more), if we consider Dell’s (of which Alienware is a part of) net income from 2013 hovering around 702 million dollars. Of course, Dell does a lot of other stuff, but Alienware was already a 160 million USD business (revenues) when it was bought by Dell in 2006, and it’s been rumored to have grown massively since then. Even for Alienware this may not be some significant sales (and on Amazon the Alienware Alpha has about 10 times more comments, suggesting much stronger sales than the Steam Machine version).

In any case, several dozens of thousands units is clearly not hitting any mark no matter how you look at it. If we consider the install base of 125 million gamers on Steam, this means less than 2 folks out of 10 000 bought a Steam Machine. That’s a very, very low percentage, among your existing client base. This also means that the Linux community (if we assume 1% of 125 million) is at about 1.25 million Linux gamers potentially – so any change coming from the Steam Machines’ shipped units is NOT going to register, even if they had any user survey appearing in their SteamOS interface.

There’s no point re-hashing the obvious (there was no marketing, no budget spent in advertising, no branding whatsoever, no framing to explain why such machines are great, no categorization of Steam Machines based on their specs, machines were overpriced, etc…), but while I am fairly confident SteamOS is a good idea (and we will see very soon when I disclose the latest survey results from early 2016, its share is constantly growing), the Steam Machines by themselves may not be ready for the market right now. I still think they will make a lot more sense down the road, when hardware becomes cheaper and much more powerful, but there’s still many features missing in SteamOS (we still don’t have any video service, and absolutely no VR support) and Vulkan is just starting to gain a little traction. Problem is, will manufacturers be willing to keep Steam Machines in their catalog 2, 3 years later, based on the pitiful sales of Year 1 ? They may need a lot of convincing to continue supporting that initiative down the road.

Still, while articles about the platform are overly negative, I agree at least with the feeling. Valve did not go a good job at launching these machines, there were a million ways they could have done things better, and let’s just hope they are in for the long term and will learn from their mistakes, instead of just pretending nothing happened.

At BoilingSteam, we strongly dislike ads and that is why you won't find any during your visit. If you like what we do, please consider supporting us on Patreon. You can also sign up to our newsletter (No Spam!), or our RSS feed or even Twitter.

Related Posts



  1. Keep in mind Valve is developing a video streaming service of their own. I don’t know if it’s compatible, and everyone wants third party services instead. That was a glaring problem….

  2. It’s funny how the real reasons steam machines suck are rarely a focus on articles like this – sure they *might get mention in passing.

    Critical Success Factors:
    – Can play all (not some, ALL) PC games; no EA? no Id? no 2K? no Blizzard? no Rockstar? etc
    – Can play multi player split screen (also an issue on Windows PC) like ALL console game counterparts
    – and that’s it? yep – THAT’S IT. PERIOD.

    Until Steam focuses on this, gamers will continue to NOT buy linux steam machines, and they will continue to jump ship to consoles – for those two and ONLY two reasons above. I know it’s easier said than done, but not doing anything isn’t helping. Fix these two things and sales will speak for themselves and steam becomes a household name.

    Just my two coppers.


  3. To get the SteamOS userbase to critical mass Valve needs to compel the Windows Steam userbase into dual booting. How? They need (timed) exclusives, whether that be games or SteamOS features. Halflife3 for starters….L4d3, portal3, dota3, cs2, tf3, the list goes on.
    Steammachines are a nice initiative and certainly help Valve to reach out ‘beyond’ the current steam audience but they won’t help with establishing a solid userbase for SteamOS. Which is what they really want if they want to attract publishers and developers to the platform.

    100% compatibility of the windows library is unlikely to happen even if Valve is currently doing a great job. Imo securing new titles and popular old ones should be sufficient for attracting gamers. Publishers could always release older titles at a later point in time.

    • Exclusives is not in the spirit of what Valve wants to do. Steam is multi-platform, they have no interest in making one platform more interesting than the others artificially (they already said they won’t do exclusives for SteamOS).

      Steam can grow on all platforms at the same time. For Valve what matters (right now) is to expand their userbase, not to transfer one part of it to the other side – because that’s not growing, it’s just displacing users.

      Plus, since the Windows catalog is bigger, displacing Windows gamers to SteamOS may result in less sales for Valve, since the choice of games is anyway more limited on SteamOS. Less choice -> less sales. It would be negative in terms of cashflow.

      • What you say makes a lot of sense , especially in the long term but…….
        It will come at a price. The strategy you describe may work for Valve, for hw manufacturers it doesn’t cut it as they need to make money back. In the short term Steam machines will fizzle out and oem’s will abandon Valve due to lack of consumer interest. Steam machines doa indeed.

        I’m convinced Valve will only play the exclusive card if they have no other choice. Perhaps that time may come when ms decides they want control of windows gaming back. That might be sooner than expected as ms is well on course merging xbox and windows gaming via UWP and xboxLive. If MS succeeds they will become The biggest player in gaming. Only thing standing in their way is Valve and Steam for windows.

        • Valve’s problem is that they aren’t willing to take much risk with their steam machine or OS. Making anything exclusive or subsidizing the steam machines would mean spending more (before we even consider their philosophy). Console market is a costly and risky sector to break into and that’s why you don’t see many companies trying to get in like they do with smartphones and tablets.

  4. Dell just released (today) a refreshed Alienware Alpha (“r2”), and thus a refreshed Alienware Steam Machine. Besides an update to the available processors (updated from Haswell to Skylake), more graphics options are now available than the ho-hum “custom” NVIDIA GeForce GTX 860M that the Alienware Alpha and Steam Machine were sporting. While the Alpha adds both the AMD Radeon R9 M470X (2GB GDDR5 configuration, not the 4GB standard) and the NVIDIA GeForce GTX 960 into the mix, given the state of AMD graphics performance on Linux, the Steam Machine adds only the GeForce GTX 960. Not a big deal, as 1) per ( and Game Debate (, the Radeon R9 M470X may be less performant than the existing GeForce GTX 860M, and 2) obviously the desktop-class GeForce GTX 960 would be preferable to either the GeForce GTX 860M or Radeon R9 M470X mobile GPUs anyway. Another difference between the Alienware Alpha and the Alienware Steam Machine is that the Alpha is configurable with a 7200 RPM 6Gb/s HDD and/or a M.2 PCIe SSD, while the Steam Machine has only has a 7200 RPM 6Gb/s HDD. The Alpha is also compatible with Alienware’s Graphics Amplifier. The Steam Machine details don’t include mention of an Alienware Graphics Amplifier port.

    The top-of-the-line configuration of the new Alienware Steam Machine (i7-6700T, GeForce GTX 960 with 4GB GDD5, 8GB DDR4-2133, 1TB 7200 RPM 6Gb/s HDD) should now outclass the first iteration of the ZOTAC NEN (i5-6400T, likely a GeForce GTX 970M rebadged as a GeForce GTX 960 with 3GB GDDR5, 8GB DDR3L-1600, 1TB 7200 RPM HDD) at roughly the same price point (≈ 999 USD). Too bad Valve doesn’t subsidize the cost of any of these Steam Machines. And the new Alienware Steam Machine is still not quite ideal for VR, not that the HTC Vive and Oculus Rift support Linux yet anyway.

    Dell is the “winner” of the Steam Machine wars thus far, and it was positioned to be from the start (although the Steam Link made all Steam Machines an even less attractive buy than they already were). Of the Steam Machines, the Alienware Steam Machine received the most promotion from Valve, and it also received the product placement at brick-and-mortar stores (GAME, GameStop and EB Games), which, unless I’m mistaken, is a deal that was primarily arranged by Valve, not Dell. Of the existing Steam Machine manufacturers, Dell also has the most resources behind it. The only heavyweight which could have possibly matched Dell is ASUS, and it pulled out of the Steam Machine debacle. It’s a shame, although understandable, that the Steam Machine product presentation has been somewhat Alienware-centric, as I think the ZOTAC NEN shows the most promise as a console.

    I think what the future holds for Steam Machines is exactly the approach Dell is taking. The Windows equivalents of the Alienware Steam Machine (the Alpha), the MAINGEAR Drift, and the Syber Steam Machine (the Vapor) have to be outselling the Steam Machine configurations. Going forward, I think Windows configurations of SFFs will receive greater focus, but some manufacturers may offer Steam Machine configurations of these SFFs in addition. What I don’t think we are likely to see going forward are official Steam Machines which don’t have comparable Windows equivalents (i.e. form factors devoted solely to SteamOS), unless Valve finally unifies the platform. I also think the number of Steam Machine manufacturers will whittle down further. We started with 15, we are now down to 6. There may be new additions to the list of manufacturers, but, if so, I think they will probably be offset by manufacturers who stop offering Steam Machines. I think Dell will probably stay in the game as long as they have competition in that space. I hope ZOTAC sticks with it, and if Valve were to pick one hardware partner with whom to create a uniform, Valve-branded Steam Machine (which is what they should have done in the first place), of the existing manufacturers, ZOTAC would be my choice.

Leave a Reply

Your email address will not be published. Required fields are marked *