Cloud computing for video games… true or not?

photo credits: zdnet blog
photo credits: zdnet blog

Dear all,

A few weeks ago, I posted an entry on this blog about Cloud Computing for video games, including a poll. The poll results show that 40% of those who voted considered this as “definitely interesting”, while 29% of voters thought “I don’t think it can work, at least for the next few years”.

I personally agree with 29% of you then!

Let me explain. Cloud computing is nothing new, the term “Software as a Service” emerged in the very late 1990s (source). Actually these initiatives from Onlive and now Gaikai (announced just a few days after Onlive), are a pure transposition of this model applied to the video game market. As you may know, PC game sales has been declining for years, among others because of software piracy, and game developers and publishers have been concentrating on building consoles games to allow a more stable environment than Windows (with some many possible configurations, drivers, video cards installed etc.) and access people in their living room.

Today, it’s a perfect time to make announcements like the one we saw at the Game Developer Conference. Game developers and publishers are trying to find alternatives to the traditional brick and mortar sales channels to concentrate further on online delivery. It’s already successful with Xbox Live Arcade and growing with rivals Nintendo and Sony respective DSWare, WiiWare and PlayStation Networks.

I don’t deny at least part of the technology promised by these actors is working, though I think it will generate huge technical issues when deployed even in beta, but I think such shift will take years and years to become more than a drop in the game developer and publisher revenue share.

What do you think?


  • Rahul V Suryadevara

    Hello Virgle,

    A few weeks back, I was wondering when on-demand product development will effectively use cloud technology. However, for gaming, I agree with the 40%. One of the key differentiator is the size of the existing consumer society. Gaming is evolving to be the next big media format with active man-in-loop (interactivity) as key USP. I personally feel cloud technology is the next tipping point for gaming revolution.
    The demand for more and more realism in gaming is going to make cloud computing a critical element in sophisticated game delivery. I personally think next 3 years are going to be the strategic turning point for this technology and we will start to see the main stream commercial products being delivered to public using cloud technology by the end of 3 rd year. (end of 2012).
    Imagine the possibilities of applying this technology to behavior pattern games like sim xxx and ability to introduce variety of more accurate physics based models dynamically in to the internet games. Cloud can accelerate user reaction feedback to game developers and help to modify and deploy the new models in shorter time.
    I will not be surprised if automotive companies start validating their new models with accurate physics based behavior models in the game world first. Everyone can be a James Bond in game world.

    Am I dreaming? ………. (Revalidation is involuntary for any engineer, I guess)

  • I think cloud computing and online will be next revolution in gaming. Our life becomes more social and games will reflect it. Oleg.

  • Hi Rahul,

    I agree with you for some of your points, but I must emphasize the *still* growing power of PCs and the next generation of consoles will counter the need of the need of cloud technology. The next generation of consoles is now a bit farther away due to gameplay innovation being added (through the motion controllers) and the current economic crisis (which may have forced the latter, which is great then!).

    To Oleg,

    The online part is even more right indeed. Being able to play in a browser, while chatting with friends eventually directly within Facebook or so is part of the future casual gaming experience. The monetization remains a big problem to solve though, as currently browser based games = FREE to consumers most of the time.

    Personally, I think the number of passive players will grow, thanks to the development of the so called “spectator mode”. Watching a virtual race on your TV in real time, being able to simply switch cameras, get your own UI etc. will develop fast as well. I think I should actually plan a post about that in September 🙂

  • KDI

    There’s no doubt online functionalities will be more and more central for the gaming industry. However, let’s make the difference between the online functionalities per se and the technological choices made in order to see them happen. Allow me to share with you some of the points we discussed with Virgile… And bear with me, please. This comment is quite long!

    Back when I was at school we used to have access to most of our software through terminals. The program would run on servers and the users interacted with light clients. This was quite usual at that time and in many companies too. I found this was convenient for software types you couldn’t find on PC and for software that needed a lot of processing power and long computation time.

    Then, there was the great Client-Is-King revolution.

    The IT gave us access to more and more PCs and less and less “light terminals”. And I thought it was for the best: I could still connect to the servers from the PCs and concerning the software run on the client side, I didn’t have any latency problems anymore and I wasn’t impacted by the inevitable bugs/crashes/memory leaks potentially triggered by the runtime of other users. From a user experience standpoint, running most of the software on the client end seemed like a great improvement for my usual applications let alone highly interactive real time software like CAD or even games I programmed myself that needed the best feedback performance I could get. And the future seemed to be even greater as I was assured to have better machines as long as enough money was invested (Moore Law). So from a professional point of view, my situation was better. From a personal point of view, I had to deal with my PC and the poorly designed installation programs…had to buy the retail software, whether I was going to use it for a long time or not, paying a sometimes hefty price for the hope of satisfaction. I didn’t have another choice at the time anyway.

    Now, some people (Gaikai and Onlive resp.) come and tell me that, at least for my games, I may have the choice! All the processing will be done by a server and I’ll have a light client at home. I’ll be able to benefit again from the wonders of heavy server computation and light clients, and that this should be a great source of amazement and satisfaction. Well… I have to wonder…

    This seems like a Server-Is-King revolution to me. A step backward in history? Clearly, this revolution can’t be for the better in terms of technological efficiency:
    1) D. Perry basically says that if the performance is bad, he’ll put a server next to your house which is not very far from having it inside your house…
    2) If this was the case, the professional world would probably have already done it.
    Dave Perry and Onlive People are no newbie in the industry, so if such a revolution can’t be beneficial for the professional world, then there should be some other benefits a consumer could have at home but not at work…

    Diversity/variety could be an obvious one. At work, the programs I use don’t change very much over time. There are upgrades, but basically, I always use one to type my texts, one to program, one to communicate etc… I don’t expect diversity and discovery in my user experiences at work, I run away from them for the sake of efficiency and to reduce the cost for training. At home, this is quite the opposite when it comes to entertainment. If I don’t get enough diversity (in a given personal frame), I’m bored. So having access to a wide selection of online entertainment experiences makes sense for a consumer. But diversity isn’t enough to justify the switch to computation on servers: VOD (video on demand) and app stores already provide the first without the latter.

    So why all this fuss about Entertainment Server computation when Moore Law allows us to have super powerful machines in the palm of our hands? The benefit should be obvious, for the consumer to buy into that.

    The answer may not really exist yet. My guess is, only a new range of multi-user ad hoc applications could be the key differentiators, which basically boils down to the so popular “Content is King” motto.

    OnLive and Gaikai people will have to come up with experiences that use the centralized computing to their advantage in order to convince the consumer. The so-called spectator mode could be one of them. But a 100% passive spectator mode doesn’t need centralized computing. It only needs centralized data storage as one could watch tomorrow what is happening now in my game instance without even noticing it’s not real time feed. Nintendo (Super smash bros brawl); Sony (WipeOut HD, Killzone II) and Microsoft (Xbox Live) already provide that. However a mostly passive mode, where the spectator could occasionally and lazily try and influence the courses of the many games he’s simultaneously watching, would benefit from server computation and provide and interesting twist on the already existing spectator modes. If such Onlive/Gaikai systems proposed their own avatars, and games using them, they could imagine modes where your avatar drops out of one game to be thrown unexpectedly into another instance of another game, based on your tastes, your friend list… Mixing game world instances with different proportions would be something really interesting to explore: simultaneously, one after the other etc… Plus this would add tremendous replay value and could break the linearity (but also the consistency perhaps…) of some game genres…

    I agree with Virgile when he says the announcement is cleverly timed though, even if the true revolution it could be is not yet mouthed. It is early enough to get investors on board and to get the feedback of an industry which is not really in a hurry to switch to the next generation. There’s much to say about the economics here too (business model, ecosystem) but I’ll leave that for now…