Could Improbable Technology Make Massive Virtual Worlds Possible?

I have just read an article on Endgadget : Better than ‘Destiny’: Studios now make massive games in just months. The article centres around technology from a company called Improbable and this sounds very interesting indeed. The CEO of Improbable, Herman Narula, is quoted in that article as saying :

It’s about having no game server. What you’ve always seen has been game worlds where there’s one server for one region and another server for another region, these very neat lines. And the servers are actually just the same game engine that’s running on your computer. … But what if you didn’t do that? What if, instead of that, actually you had thousands of tiny, very limited processors — call them mini servers, like a swarm of insects? And what if, instead of having boundaries, actually they all moved around many, many times a second, migrating to deal with simulation in a particular area? And they’re all able to work together to model a world much bigger than any one of them could understand.

Whereas the focus in the articles I’ve read on the Improbable technology have been about games, I’m pretty sure these concepts could also apply to virtual worlds. Indeed one of Improbable’s employees, Dave Hillier, once worked on Second Life, so the company have someone with knowledge of virtual worlds on the team.

The Endgadget article explains a little bit more about what Improbable is and isn’t :

Improbable isn’t just a series of servers. It’s cloud-based, but it’s not cloud rendering; it’s almost an operating system. It follows in the fresh footsteps of other studios crafting large worlds with just a few people.

What they seem to be talking about is a potentially huge world that reacts in a permanent fashion to player actions, talking about the game Worlds Adift, the article states :

the game world reacts in permanent, persistent ways to players’ movements. Build an airship and drop a boulder overboard, knocking down trees and crushing players below. Come back to that same bit of land months later, and those trees will still be knocked over, perhaps with other plants growing around them, or with other players harvesting them for resources. Real persistence, real in-game consequences to physical actions. This impacts not only mechanical moments, but also the story that Worlds Adrift tells.

Those are the sort of concepts that make me ponder whether this sort of technology would be suitable for a virtual world environment, because in virtual worlds, especially user content generated worlds, changes are largely persistent.

Dean Hall, who created the game DayZ is a big fan of Improbable, writing on a Tumblr post he says :

Last year I met a company called Improbable. My first meeting with Herman Narula, the CEO from Improbable, was one of the most surreal I ever had. The technology I had always wanted and tried to make was finally here. DayZ was born out of my aborted attempts to make a database architecture to support my wild mass multiplayer ideas. But now, I didn’t need a ten year plan to make my grand visions of multiplayer come true. I could do it now.

Dean, who recently climbed Mount Everest, also writes :

Working on my first improbable game is the most exhilarating thing I have ever done.

I don’t know enough about the technology behind Improbable to know whether it would be suitable for a virtual world environment, it sounds good, but how it delivers is a different matter. However the old methods of how virtual worlds have been structured are going to change.

High Fidelity’s model is very different to previous models as High Fidelity plan to harness distributed power from computers all around the world to deliver their virtual world. Improbable use a different model. Both models are designed to deliver massive environments of a kind we haven’t previously seen.

I’m sure there will be more technologies emerging in this field, it’s an interesting time.

Improbable website : http://improbable.io/

Tweet