Xbox One - Head in the Cloud

Microsoft has a plan for the future of gaming - and it's one that leaves me a little worried.

Xbox Live Compute is a project that will see Microsoft's Azure cloud servers used to bring all sorts of computational benefits to Xbox One games - from dedicated game servers to more intelligent AI to improved rendering techniques. There's a big article about it at Gamesindustry International in which Microsofts John Bruno expounds on all the great things cloud computing can bring to Xbox One.

And it sounds great on paper. The idea that huge Microsoft servers can handle aspects of a game from the other side of the world and I can see the benefits in my living one is extremely cool. It could even make the hardware specification arms race irrelevant in the future.

It also sounds like something that won't work. Not in the foreseeable future, at least. Why? Because however powerful Microsoft's Azure servers are, however efficient they make their net code, this bounty of cloud-processed power is still going to have to be pumped into people's Xboxes by actual broadband that people actually use

I currently live in a small English village, and can comfortably say that Xbox Live Compute wouldn't work in my house. My 360's relationship with my wifi is one of trembling knees and trembling connection reliability; the console will fall offline at least twice an hour and will usually not connect at all during whatever nebulous hours TalkTalk define as peak times.

This may not be much of surprise - after all, I live in one of the many areas of the UK yet to be blessed with fibre optic cable and its heady promise of High Speed Broadband For All. Logic would dictate that if I wait a couple of year for my local council to get its act together then I'll be golden, happily surfing the Azure wifi to Microsoft's gaming utopia in the cloud.

But the thing is, before I lived here, I lived in a big city and owned a cable broadband package that boasted a 30MBPS connection, and that couldn't hold a reliable connection 100% of the time. We had regular dropouts, regular bottlenecks and ruthless peak time throttling that rendered our wonderful Virgin Media package pretty much useless for stretches of time.

There are just too many things that affect broadband delivery for there to be any such thing as truly reliable broadband. Thinkbroadband.com features some wonderfully optimistic infographics about how soon the entire UK can expect to be connected to high speed broadband, but there's 'high speed broadband' and then there's 'reliably useful high speed broadband.' Or more to the point there isn't.

It doesn't matter how many supposed megabytes we're able to download per second - as long as our broadband is provided by companies who offer limited usage packages, nebulously defined bandwidth caps and peak time throttling, the idea of truly being 'always online' is a pipe dream.

And this makes Microsoft's vision for Xbox Live Compute a tricky one for me to get behind. This is a different beast to the much-loathed 'always online' server check-in that was originally pitched at the Xbox One unveiling. As unpalatable as that sounded, even my reliability-challenged connection could manage to ping a server once in 24 hours.

But when actual in-game content is being streamed live over the net, lots of tricky questions are raised. What happens if my internet drops connection during my new driving game in which the enemy AI is cloud computed? Do the opposing racers suddenly become incredibly stupid and drive into walls?

Even more concerning, what happens if parts of my game's graphics are being rendered on Microsoft's servers? If I drop connection, are things going to disappear, or am I going to be left playing some shonky low-res version of the game?

It's one thing to use the Azure servers for dedicated multiplayer games that already have a stable internet connection as the lowest barrier to entry. But when your internet connection has to bear the weight of world generation, game AI and even graphics, we're left with a stickier proposition. What happens to those of us who can't stay reliably connected for the duration of the game?

This may not be a problem fifteen or twenty years from now. But in the next five years, when Xbox One business will likely prove to be most viable, there's no way the broadband infrastructure in this country will catch up to Microsoft's dream of an always-connected future. And from what I've read of comments surrounding the original always-on Xbox One proposal, the situation in America isn't a whole lot better.

To me this is emblematic of Microsoft's entire Xbox One strategy; pushing for too much, too soon. The company seems desperate to dive into a purely digital future; perhaps it's anxiety over their market share in personal electronics being rapidly devoured by powerful rivals. Perhaps Microsoft needs to build a cloud-powered experience into everything because if they don't then Apple or Google might do it first. But in doing so, they run the risk of creating products that a large chunk of their potential user-base won't be able to actually use.

There will come a time when we all have genuinely reliable high-speed broadband connections. There will come a time when people will happily accept the idea of an always-online games console. There will come a time when most of our digital content exists on  cloud server.

But that time isn't now. It's probably not five or ten years from now, either.

It's possible Microsoft have this all figured out. It's more than likely that Xbox Live Compute won't be nearly so influential as that article seems to suggest, and as service for providing dedicated servers and improved multiplayer capabilities it seems like a solid idea.

It's also possible that Microsoft is more than happy to target only the type of people who already do have all the requirements needed to benefit from Compute, and hang everyone else out to dry, the way they initially seemed to be willing to do to broadband-less gamers with the Xbox One unveiling.

But assuming that Microsoft is a company that follows an internal strategy that isn't insane, and actually wants to appeal to the widest consumer base possible, then betting so heavily on cloud computing this early in the technology's lifecycle - especially given widely-publicised failures like that of OnLive - seems to be running a big risk of fragmenting an audience that already seems less than entirely enthused with the Xbox One vision.

Time will tell whether cloud computing really is the future of gaming, and whether Microsoft's vision for Xbox Live Compute really will be game-changer it claims. But while broadband access is still primarily controlled by local councils with no interest in high-speed internet and internet providers that seem to have no incentive to offer competitive or compelling packages, I will be eternally wary of anything that requires that I be online.

Because in reality, I would love to be always online. I just don't think it's going to happen any time soon.

Tom Battey is an author and person who sometimes writes about videogames. He writes at tombattey.com and does the Twitter thing @tombattey.