If you haven't seen the video showcasing OnLive's online gaming platform from this year's Game Developers Conference, you should check it out here.
OnLive's product is simple: deliver high-end gaming experience of multi-platform games (PC, XBox 360, PS3) through a browser. Supposedly, with very little porting, game developers can adapt their game to run on the OnLive server platform.
Users connect a controller to their PC, open up a browser, and faster than you can saw "Lara Croft is hotter than the sun," they're playing real games via the internet. All of the controller's movements are sent via your high-bandwidth internet connection (at least 1.5MB for SD, 5 MB for HD) to one of the servers in OnLive's farms, where the game is actually running. Only the video itself is sent back to your screen -- you don't need a GPU, you don't need a high end system at all.
For people who want to game on their TV without a PC, OnLive is manufacturing a small box (the size of paperback book) that plugs into your home network and can accept wired (USB)or wireless controllers.
In addition to offering games from multiple platforms without investing in lots of expensive hardware, OnLive claims to have some value add on top of the games themselves: improved social networking, the ability to save "brag clips" of your best moves, and the ability to watch other people play games are all built in.
It seems that latency would be a huge issue, even on those high bandwidth connections -- they claim that it's imperceptible, but only real game play will tell.
If their product does everything that they say it does, though -- this could be the "killer app" that cloud computing has been waiting for. This could quickly turn a multibillion market on its ear. Why invest hundreds of dollars in a console when there is an option that requires none (and could potentially play more games)? Why invest $60 per title in games?
While some industries (say, enterprise software) may have difficulty convincing customers to move data, try something new, and pay-per-use, the video game market will have no such hurdles. They're marketing to a generation who has never purchased a CD, who use more SaaS in the cloud, and who would love to avoid the sunk cost that a console represents (I myself have a PS1 and a PS2 gathering dust downstairs).
OnLive has that rare opportunity to be groundbreaking in two industries (gaming and cloud computing) simultaneously.
And if it is the killer app, there is a strange side effect: while many people have been assuming that servers would be the first industry killed by the move to cloud computing, it would be the console manufacturers (Nintendo, Sony, and Microsoft) who get affected most.
Update 3/25/2009 4:16 - added the last paragraph
Thursday, March 26, 2009
Friday, March 13, 2009
I hadn't actually forgotten that I have a blog...but from the date since my last post, it certainly looks like I have.
Certainly 2008 was a tough year, and I ended it by entering that fraternity known as "fatherhood" -- so I've spent less time blogging than I should.
I have still been following the grid and cloud spaces quite closely, though, and plan to start crystallizing more of my thoughts here.
In the meantime, some Digipede-in-the-news: Penny Crosman at Wall Street & Technology wrote an article called Adapting Legacy Applications to Multicore Servers that featured Digipede very prominently.
I was glad to see it, because it's one of the benefits we've been touting for quite a while now: many enterprises have a decade (or more!) of legacy code that they run, and any multi-core/multi-machine strategy (whether it is internal to their data center or external in a cloud) absolutely has to address the issue of how to adapt that code to take advantage of newer hardware.
A Digipede customer is quoted in the article quite extensively, but my favorite quote by far is this one:
...staff have become almost obsessed with throwing applications on the grid because it's so easy to do.
Posted by Dan Ciruli at 4:50 PM