The Age of Enlightenment

Geomerics has been going for about three and a half years now. I founded it as a spin out from Cambridge University - we had a number of academics who were interested in commercialising some of their work, and I put together a business plan. We got together an initial investment of about GBP 500,000 from a Guildford-based angel fund. I left the Physics Department at that point, set the company up, and we put together some demos.

Then we went out and talked to as many developers as we could, and came back with a very clear impression of where the technical challenges were for this console cycle. People identified the three hot spots as physics, animation and lighting - but most people were fairly sure that physics would be solved by Havok, maybe Ageia, animation was probably going to be okay with NaturalMotion and Autodesk, but it wasn't clear who was going to pick up lightning... particularly with Renderware disappearing.

So that seemed to be the market to go after, and we focused all our efforts on developing what we thought would be the killer app for lighting - which became Enlighten. We thought there was a part of lighting that could be made real-time in this console cycle that people didn't realise. There are some things you can do in real-time and some things you can't, it depends on how you structure your data and algorithms - but we felt that radiosity could be computed in real-time, and could be a potentially enormous saving in effort for developers, and lead to new interesting gameplay and much better control over the look and feel of a game.

We've been developing it for two and a half years now - it had its first launch at GDC last year. Since then we've been working with the early adopters, getting the technology around and about, and started announcing deals towards the end of last year. We're going back to GDC this year with the second iteration, which is taking the basic functionality, but now extending it to the scale where we can do entire cities. We can pretty much cover all game genres now, and it's also now integrated into the Unreal Engine, which is looking like it should bring in quite a few customers for us.

We did that in a number of ways. First of all we clearly set out to try and raise the bar within the industry as far as lighting was concerned. There were two things that we thought were important - one was realism, so was it actually realistic, what we were seeing? Also, what areas could we improve? We could improve gameplay through creating a better mood, a better atmosphere.

Those were the things we were describing to people that we could do with Enlighten, that couldn't be done before.

We've signed a number of deals, but we've only announced one so far.

Absolutely.

We've been quite surprised by the variety. Even the first one with EVE, being a PC-based MMO, wasn't what we thought our core market would be.

We've got another PC-based war game, and there's a survival horror which we thought would have been a more obvious choice for us.

We've been surprised at how different they've all been the deals that we've done, and the deals that we are going to do that we know of. Because part of our business is that we get to see the kind of games before we do the deals with people, we have to work with developers for a certain amount of time on their projects - so it's fair to say that the projects we're working with are very varied.

Standard terms are that it's a fixed fee per game.

Well, you always get excited about your own technology, but when we took it to GDC, we sat down with people, showed them the demo and they were astounded. Particularly artists, because they can see something they can immediately use, and then reality starts to set in and they can see the potential productivity savings as well - it's done in real-time, so not only can they create the world in real-time, which has never been possible before, but they'll have loads of time left. So the CTO is thinking they can either save time, or spend more time doing something even better than they thought they could do initially.

Well the savings come from the fact that Enlighten does it in real-time, so any change within what you're doing happens instantly. If you change a colour, like a carpet, or wall, you get an instant result. There's no pre-baking going on, which can take anything between 8 hours and 24 hours, depending on which game it is.

We reckon that on a high-end title you're probably saving between two and four man-years of art time, and a not dissimilar amount of coder time as well. The potential savings can be around GBP 250,000, so we reckon games will actually make money by spending on Enlighten.

We calculated that on a general development timeline, Enlighten would shorten it by about two months - it depends on the size of the team, and the size of the project, but it's in that region.

For content generation it just runs on PCs. For in-game on the PlayStation 3 it runs on one SPU, it really doesn't touch the sides, there's always that resource spare. And on the Xbox 360 it runs on a separate thread, and pretty much every developer we've spoken to has had a spare thread kicking around that they haven't been able to utilise effectively.

It actually hasn't been a big problem, finding the runtime resource for it.

Yes, it does - I think it helps a great deal. I think our original goal was to do that - to take Enlighten, and other products over time, which were traditionally some of the more inherent problems that the industry's had, and solve them. So we'd become a specialist team for areas that were difficult.

I think also, the way the market's gone, it only speeds up our business plan - but ultimately what we're about is to allow game developers and publishers to do what they do best, which is just create games. We want to supply them with solutions that have historically been difficult, but we can concentrate on them and solve them, and help increase the gameplay experience in the market place - which we think is incredibly key during these difficult time.

I think every industry is looking at costs. Even the games industry has to be very aware of what they're spending on games. We're fortunate in the fact that the industry is still very buoyant, and still growing at a tremendous rate - but we have to be aware that this probably will not continue at the extent that it has. I'm sure developers and publishers are looking at the overall costs, and they're looking at areas they can save - and Enlighten would help them.

But more importantly they have to look at areas they can use to increase the gaming experience going forward. I think in this difficult year the gaming experience has to be increased dramatically, we have to continue to persuade people to buy games, which means we probably have to speed up our technology and deliver bigger and better things on a more regular basis.

Business is very strong for us right now, and I think it's only going to get better - that's where we see ourselves.

We have a number of ideas which we have proven out, but that doesn't necessarily mean that's what our customers want. So as we continue to grow the company through Enlighten, our intention is to listen and work with publishers and developers to identify the key areas that they would like.

So we're in the throes of that - and therefore it's very difficult to say we'll have something specific in January 2010. It's not quite as solid as that - it's a moving, fluid thing.

We have to keep close relationships with all the hardware vendors, whether it's Sony, Microsoft, Intel or Nvidia, and we have to have a pretty broad research base, which we have through our links back into Cambridge University, where we can secure IP very easily.

And also, we're on a GBP 500,000 grant from the Technology Strategy Board working with University College London, which is securing our graphics research for the next three years.