Cloud Cover Clearing: OnLive Shutting Down

I was surprised to hear that OnLive is shutting shut down, mostly because I’d forgotten it was still running. The cloud gaming service launched in June 2010 to a chorus of people asking “Wait, so I’d pay to have PC games look and control worse than they would on my PC?” It was a solution looking for a problem, and decided its problem was that folks who like big fancy PC games wouldn’t have PCs that could run them.

OnLive have announced that they’ll close the service on April 30th, taking all users’ streaming games and data with them.

OnLive support’s FAQ about the shutdown explains:

“After April 30, 2015, our data centers will shut down and the service will be offline. All accounts will be closed, and all data deleted including game save data, achievements, and credit card data will be deleted. If you purchased a Steam game from OnLive, that game will still be available on Steam. No refunds will be available for any game purchases, hardware purchases, or subscriptions.”

Cloud gaming, to briefly explain, runs games on remote servers. Player control inputs are fed to that system over the Internet, and they watch a video stream of the game being played sent back from that system. It means that you could play demanding games on PCs and tablets capable of watching HD video, but games looked worse and felt laggier than if they were played on a PC. OnLive had an audience, but a small one, and it stumbled out the gate with subscription fees and an unexciting library of games.

“Overcoming the perception of being dead has been one of the unanticipated challenges of the turnaround,” OnLive say in the official announcement. The service almost went under in 2012, and its comeback was unspectacular. “In fact many of the recent articles that mention OnLive refer to it as ‘defunct’ or something similar.”

OnLive had hoped they’d find a buyer who wanted to continue the service, but couldn’t. “Most of the companies that declined to acquire us did so due to the perception that they did not know how far off in the future cloud gaming would be. Predictions that cloud gaming will only be far off in the future are self-fulfilling prophesies.”

Instead, they’re selling an unspecified chunk of their assets to Sony, probably tech and patents, and shutting down. Sony already dabble in streaming games by letting folks play PS4 games through Vita handhelds and other devices, and have plans for more OnLive-y cloud gaming.


  1. Axyl says:

    I genuinely thought OnLive had shutdown ages ago.

    • Shadow says:

      Same. I thought it never took off, but I guess that’s sort of true. Just didn’t expect it to languish for so long.

      Until Internet services achieve virtually realtime, LAN-like speeds, cloud gaming cannot seriously compete. And even then, physical distance to the game server would remain an ineludible obstacle and make the whole prospect unfeasible unless there’s one such server within a few hundred kilometres of your computer.

      • ShadowTiger says:

        I think in countries like Japan and Korea they have much less problems with latency and distance to servers. It might be more viable in those admittedly smaller markets.

        • battles_atlas says:

          And yet in such a small, homogenous market almost all the potential efficiencies of cloud gaming are lost. The ideal would be a huge marketplace spread over lots of timezones. That way economies of scale are huge, and demand is distributed over the 24hrs. Then the economics seem to make more sense.

          I never really understood how this was supposed to work in the real world of network constraints.

    • Shuck says:

      They more or less did – in 2012 the original company collapsed, they laid off all of their employees and the technology got bought up (for a pittance) by another company that continued operations. That company didn’t have any better luck with it, apparently. Unsurprising given the niche appeal and problems.
      I’ve always been skeptical and somewhat leery of the service, but the game industry loved the idea – this was either an additional revenue stream or it was seen as eventually replacing the retail model entirely as a way of totally eliminating piracy.

    • Montavious says:

      I totally forgot they existed too. Couldn’t have helped not to be advertising. I didn’t think it would even last this long.

  2. Premium User Badge

    Aerothorn says:

    Yes, Sony is buying a bunch of patents. Which is kind of scary – given American patent law, the combination of these patents and the ones they got from Gaikai will basically ensure that they’re a cloud gaming monopoly for the forseeable future.

    • gunny1993 says:

      Fortunately, given us internet speed and quality, no one over there will be able to capitalize on such a monopoly.

    • Horg says:

      In 5 years the patents for the OnLive hardware and software wont be worth the paper they are printed on. In 10 years they will be in a museum. They don’t have a patent on the concept of streaming games, so if someone else wants in on the market they just have to create some superior hardware.

    • airmikee says:

      Considering that OnLive has already failed once because of a lack of paying subscribers, that means Sony is likely to have a monopoly on something no one really cares about.

      • Premium User Badge

        Aerothorn says:

        Different marketplaces. I think Playstation Now is already moderately succesful, and the ability to have your PS4 be a one-stop-emulation-ship is a very powerful attraction.

  3. Kollega says:

    And yet, in 2008 or so I read an article on Absolute Games, formerly the biggest Russian gaming portal, that prophetized that cloud gaming was “THE FUTURE” for PCs and that in 2015 the idea of a high-powered gaming rig would be going the way of the dinosaur. I think the guy who wrote the article just played too much Crysis and assumed that all PC games in the future would require such hardware. And in fact, this also reminds me of evaluation of the original Modern Warfare by Igromania, a big Russian gaming magazine, that they were running at about the same time: “The apex of scripted shooters’ evolution. The genre is following Crysis now, but missing Modern Warfare would be a crime.”

    Moral of the story: don’t make predictions and say they’re definitely going to come true if you don’t count on looking like a fool a few years down the line.

  4. Smashbox says:

    So cool of Sony to buy them and then shut them down.

    Then again, I tried it once and it wasn’t very good. Did anyone use it?

    • April March says:

      I tried it and, considering that my internet at the time was horrible, I am in Brazil and therefore half a world away from their servers, and their own software was shouting at me that I wouldn’t be able to have a good experience, I did have a good experience. I didn’t notice any control lag and while the graphics were a bit blurry I never cared much for that anyway. I didn’t actually buy anything there but if their catalogue was more comprehensive I’d find them a much better alternative for playing AAA games than having to have a PC that can actually run them.

    • Villephox says:

      Sony didn’t buy the company. The company went under on its own, and has sold stuff to Sony.

    • Don Reba says:

      So cool of Sony to buy them and then shut them down.
      Then again, I tried it once and it wasn’t very good.</blockquote.
      A friend of mine tried it, but was then voted Worst Company in America.

    • clweeks says:

      It was, in my experience, an amazingly low-latency connection. Remarkable. Here in the midwest US, I could only rarely detect any control-lag.

      Still and all, what’s the point? The only serious advantage they offered was try-before-you-buy.

    • JonWood says:

      I used it shortly after launch to play some games on an aging laptop while staying with the in laws over Christmas, and was actually quite impressed. On the whole it was pretty smooth, at least until I hit the point where Virgin throttled their internet connection for streaming all that HD video.

    • BlueTemplar says:

      I tried it a few years ago and was amazed to the graphical quality of games you could run on a pretty average smartphone. The input lag was very problematic though for the racing game I tried… but my Wifi connection might have been the cause of that.

  5. LionsPhil says:

    It’s about time. Everything you needed to know about how user-hostile cloud gaming is is in that FAQ quotation.

    Except for the lack of mod support. And the complete unsuitability for low-latency interactivity. And the aqueducts.

    At least this should stop RPS talking about how promising it looks.

    • malkav11 says:

      Yeah, I’m sorry for the people losing their jobs, and I’m sorry that people are losing things they paid money for even if they never should have paid that money in the first place. But I am not sorry at all to see OnLive go. I was desperately afraid it would take off because it’s a terrible, terrible thing for consumers and what rights they still have left. It may have promised something, but not anything I would ever want.

    • The_invalid says:

      I am very glad I only ever spent £1 on that terrible service.

  6. bovine3dom says:

    Onlive was great back when I only had a netbook at University.

    If it wasn’t so easy to pirate games, I would be grumpy about having to rebuy the ones I ‘owned’ on the service.

  7. Anthile says:

    It was never going to work with their business model. Current internet structure is at least 20 years away from something like this being viable.

    • Smashbox says:

      20 might be a BIT of an exaggeration

      • ThatFuzzyTiger says:

        No, no it’s really not. OnLive’s requirements for a decent experience meant you pretty much required Fiber grade at every hop with minimal contention. We won’t have that for a good two decades at least, longer in a lot of countries. Only places where they’ve built fresh infrastructure -very very recently- are in a position to satisfy that kind of setup.

        • airmikee says:

          Fiber requirement? *rolls eyes* Exaggerating to try and prove another exaggeration. You sound like the people that claimed OnLive would kill gaming rigs.

        • ThatFuzzyTiger says:

          Considering I know about internet traffic infrastructure and shaping (it used to be my job), and considering what people would consider “acceptable” for OnLive and therefore the kind of bandwidth and latency sensitivities involved, even Nvidia shield tends to only run 1080p if you have a wired connection to the computer.

          Steam link might achieve it wireless, but that’s -in the same house-

          OnLive was not ever going to be able to deliver that kind of experience reliably, or without suffering the kind of input lag that would make people visibly cringe when playing reaction sensitive titles. OnLive was never a threat to gaming consoles or gaming rigs. But then I’ve always been of the opinion that when consultants and salesmen use the word “Cloud”, you can safely substitute the words “Snake oil” and get more or less the same results in most cases right now.

      • TechnicalBen says:

        Local hardware will always have lower latency than remote. While not a necessity (we can play the great catalogue of older games), many want the latest and greatest.

        Which means remote servers have the same problem keeping up with new tech (say for example VR on the horizon) as local computers will (needing upgrades etc).

        Games consoles are a great example of this. Easy to remote stream and host their content, yet the companies still mainly prefer the local hardware over the streaming service as it’s such a better logistic (less worries about infrastructure) and cost (each user pays up front for the device in full).

        That is my observation anyhow, so I might be wrong on the conclusions.

        • kavika says:

          > Games consoles are a great example of this. Easy to remote stream and host their content, yet the companies still mainly prefer the local hardware over the streaming service as it’s such a better logistic (less worries about infrastructure) and cost (each user pays up front for the device in full).

          Less worries about network infrastructure, maybe. But retail infrastructure and reseller markets are extremely painful. Console gaming companies are just used to it :)

  8. Wisq says:

    OnLive never made any sense to me, from a point of view of them actually making money.

    So I gathered the idea was that, rather than having to buy and maintain fancy PC hardware, gamers could subscribe to this service and get themselves access to virtual PC hardware, where they could play said PC games.

    Since they charged by the month and not by the amount of time you used the service (I think?), I can’t imagine them making profit on the “power users” who sit there using their service constantly. So I guess they were hoping to get a lot of periodic users, who would show up now and then but otherwise leave their servers mostly unused for other people to use.

    The problem you then run into is the classic “run on the bank” server capacity problem: As soon as you release some fancy new game (or a fancy update occurs for that game), you need to have the server capacity to handle all the people who suddenly want to play. Otherwise, you’ll be turning people away and they’ll start to wonder why they have a subscription in the first place.

    This is a hard enough problem to solve when you’re making e.g. a web service that can serve thousands of requests per minute per server you add, and when you can use cloud services (e.g. Amazon AWS) that let you scale your server capacity up and down and only pay for what you use. But surely they would need specialised hardware in their own datacentre to run this (due to the video card requirement), and they would need to maintain a ton of these servers to have enough peak capacity.

    So you have to decide on your ratio of subscribers to servers, and while I know this can be more than 1-to-1, I don’t think it can be much more than that. And that’s the ratio you want to be as high as possible in order to actually make money off this thing. I’m guessing people tend to game around the same time — usually the evenings, after work/school — and that’s when you’re going to have a peak capacity problem. You also can’t double them up across the world via timezones, because the latency would become intolerable.

    Maybe they could get exclusive publishing deals with some major games. It’s a great way to prevent piracy, from the game publishers’ point of view. But that requires the service be popular enough first, or the publisher will just rightfully see it as losing money by restricting their audience. Plus, making any popular game exclusive to their service would just exacerbate the “peak load” problem.

    Given all this, how on earth did they ever expect they could actually make money with this thing? It baffled me then, and I guess it still baffles them because they’re shutting down.

    • malkav11 says:

      Their original business model was “buying” access to stream individual games while paying a subscription to maintain access. Their prices were about the same as just buying the game on Steam, often higher because Steam was better at sales. It made very very little sense from a consumer perspective. Later they went to a plan where you could rent access to a pool of titles, which is much closer to a worthwhile use for the technology and vaguely close to being sensible economically. Unfortunately, it wasn’t a great selection and I think they still wanted you to “buy” anything that wasn’t in that pool. They may have made further changes after that but it was kind of off the radar by then.

    • JonWood says:

      I don’t know if this was the case at launch, but AWS will very happily lease you machines specifically designed for streaming 3D stuff by the hour. They even have APIs to send it straight to your user’s browser.

      • LionsPhil says:

        Interesting. If they had GPU-equipped boxes at all, I’d have guessed they’d angle it for GPU compute stuff.

        Do you know someone who’s making use of that?

        • JonWood says:

          They are mostly focused on GPU compute for scaling up scientific computing clusters on demand, but they seem to have realised those noses are equally applicable to gaming. The case study I saw was Eve Online allowing you to use the character creator while the game itself installs, which I thought was a pretty clever idea.

    • BlueTemplar says:

      Live Internet broadcasting like Twitch didn’t really made sense a few year ago either – one of the reasons for Xfire relative failure (while they added the video broadcasting feature in 2008) is probably that not enough people had a reliable enough Internet connection.
      In a way, it’s the same for Steam selling games and YouTube, except that it’s not a problem if the connection is bad as you don’t need the data being downloaded as fast as its being read.

      Their idea was one of scale : a few central servers and dumb terminals is more efficient than something distributed (like the Internet). The issue, as already pointed out, is that due to the speed of light limit, you still need to have many central servers.
      As another commenter already said, it might make sense in the future if there’s enough people with reliable Internet connections, but I’d have to add another condition : powerful computers being too expensive for these people.
      On one hand this already started with full-blown gaming PC’s becoming rarer, OTOH Intel processors became extremely powerful cost-wise this decade, and Intel integrated graphic chips can actually run 3D games with decent performance now.
      Maybe with DirectX12/Mantle/Vulkan developers will push the computing power requirements of games much higher and this condition will be met. We might then see a OnLive clone re-appear…

  9. KOLUNEZ says:

    So what the hell is up with the hideous mobile ads that appear when I scroll through articles? And no way to remove them, not even subscriber? Shame on you RPS.

    • Hmm-Hmm. says:

      Try something like adblock or noscript, and block ‘Taboola’ (at least).

  10. CookPassBabtridge says:

    What’s OnLive?

    Oh hahaha that was mean sorry

  11. April March says:

    The worst thing for OnLive as a business is that it shows up on RPS. Why? Because that means it’s seen as PC gaming. It’s not. It’s console gaming that runs on your PC, so you don’t have to buy an expensive console. If you want to play games from a select library, don’t mind being unable to mod them, and like shiny graphics but don’t want to mess around with system configurations, you want to play games on a console – but OnLive does all of those things for you and has no entry cost, other than the PC you probably use to watch Netflix on.

    • Yglorba says:

      I think that this is part of the problem, though. By now the gaming market (at least the ones willing to drop money on experimental stuff like this) has pretty much sorted itself out by platform. The people who want to play the kinds of games that require a huge gaming rig or a console tend to have gaming rigs and consoles.

      The people who don’t, probably aren’t interested in those types of games; it’s no use offering them the chance to play console games on their tablet when the reason they only have a tablet in the first place is because they prefer tablet games.

      • RegisteredUser says:

        What this guy said. Console experience already existed and the PC-verse frowns upon this kind of thing vs infinite tinkering possibilities.

  12. Mman says:

    My experience with it was trying it and finding out my internet wasn’t even fast enough to run it. Apparently I’m in the minority of having internet that terrible, but I’m not surprised to find it confirmed that the infrastructure isn’t there yet.

  13. Neurotic says:

    “I was surprised to hear that OnLive is shutting shut down, mostly because I’d forgotten it was still running.”


  14. GuillaumeJ says:

    The games gallery, where you could see what games were currently played and zoom on one was quite interesting.

  15. orionite says:

    I was a “founder” member and got a lot of games for free over the first year. I also got an Onlive console, which I only used a handful of times. Living in the Bay Area meant I had very low latency and I could enjoy even FPS’s. Played a lot of Borderlands on my laptop at home. It was a pretty neat service if you had a good connection. The ability to watch others play and save some clips was convenient as well.

    The key issue IMHO is, that the market for this thing doesn’t really exist. You need a good connection, so when you’re traveling and rely on hotel or public wifi, the OnLive experience suffers substantially. When you are not travelling, there is a good chance that you have a computer or a console at home, that can run the games you’re interested in, locally. The number of people who are gamer enough to want to subscribe to a service like this, but do not have other hardware that does the job, can’t be that high. Clearly, I’m just guessing at this, but the lack of success seems to support this position.

  16. RegisteredUser says:

    This was an arguibly terrible idea, but the usual marketing spin (anything anywhere anytime yadda yadda) plus the then “zomg the cloud will revolutionize EVERYTHING” hype and fear of missing out combo made things like this pop up.

    So while this was a good thing to happen (streaming also inhibits such things as modding and genuine control over your games), it was not a huge surprise.
    What does amaze me is that people still buy consoles and games for 70$. Its basically the same thing: Worse graphics, restricted interface, less options, no mods, no way to customize your experience..and the same “ease of use” argument the streaming services tried to use goes for consoles.

    Really, why do we still have consoles that steal games away from the PC as “exclusives”?

    • BlueTemplar says:

      Probably because PC, Valve efforts notwithstanding, still has trouble replicating the “couch gaming with 3 of your friends on the same screen” experience.

      • drewski says:

        To be fair, consoles don’t really do that these days either.

        • BlueTemplar says:

          *checks when Halo 3 was released*
          *sees it was in 2007, 8 years ago*
          *feels old*

          I used my laptop as a local co-op console to great effect (with the TV as an external monitor), using, depending of the number of players, a mixture of xbox gamepad / playstation gamepad / keyboard+mouse controllers.
          We’ve played games like :
          Unreal Tournament 3 (some quirks while setting up, but works great after that)
          Sonic All-Stars Racing Transformed (not as good as I remember Mario Kart 64)
          Rayman Legends
          Mortal Kombat Komplete Edition

          Lists of local co-op games here :
          link to
          And of course :
          link to
          *sees Goat Simulator there…* *starts daydreaming…*

      • Fellhuhn says:

        I play a lot of games with my friends and/or spouse locally on our couch on our PC which is attached to our TV. There is a huge amount of local coop/versus games. Even on Steam. More than on consoles anyway.

  17. Colthor says:

    The ultimate form of DRM ultimately demonstrates exactly why DRM is awful.

  18. JohnnyPanzer says:

    Everyone remembers how hyped this st(r)eaming pile of shit was, right? It was the end of PC gaming rigs. It used to make me chuckle, but then I never heard of it again, so I just assumed it never even launched. Had no idea it even got off the ground, let alone managed to cling on to whatever pathetic life it seems to have had for several years.

    Nice how you, even with a microscope, can’t detect a single fuck given in that whole “thanks for the money, we’re deleting whatever you payed for”-FAQ. It’s such a fitting attitude for the whole concept it almost feels scripted.

  19. drewski says:

    Even if the technology had worked as advertised, and the infrastructure was as good as it needed to be, there was just never going to be a substantial market for this, I think. How many gaming obsessed people can’t shell out a couple of hundred bucks once every 8-10 years for a console? That was literally the only advantage to cloud gaming – you didn’t need hardware. The software still cost the same.

    Eventually there might be a Netflix for videogames. But it’ll need to be a lot better and more convenient than OnLive ever was if it’s going to make a penny.

    • BlueTemplar says:

      I don’t see Netflix being a success either, in the current state of things, outside of a few select countries, due to region copyright management issues (and also, for most of the world, lack of high-speed Internet connection – does Netflix support pre-downloading?). This might change as their own content becomes dominant…
      link to
      link to

  20. Cantisque says:

    I tried a few demos on it a while ago. The picture quality was inferior, the latency was noticeable and the prices weren’t competitive with regular downloaded games that provided a better experience and don’t necessarily require fast Internet connections.

    There are very few examples I can think of where this service would be interesting to anyone. Perhaps you’re living at a college and only have a netbook with you?

    But for regular home use, you’re likely to have a capable PC or at the very least a console if you have the slightest interest in gaming. A PS3 or Xbox 360 would cost very little in both hardware and software and still provide a better experience than OnLive.

    Likewise, if you are away from home, you’re not likely to have access to the high-speed fibre network required to make OnLive work properly either.

    From the very beginning, I simply couldn’t understand who would use this. Playstation Now, I understand because it’s the only choice you have if you want to play PS3 games on your PS4. If the PS4 could play PS3 games natively, no-one would use their streaming service either.