In Felix's recent Gears of War 3 review he suggested that the third iteration of the Unreal 3 engine might be on its last legs. Not according to Epic's CEO Tim Sweeney, though, who says that we probably won't see a new engine from the studio for another three years yet.
Talking to IGN, Sweeney answered a question regarding the amount of time he spends working on the new engine. 'I spend about 60 percent of my time every day doing research work that's aimed at our next generation engine and the next generation of consoles,' he replied. 'This is technology that won't see the light of day until probably around 2014, but focusing on that horizon enables me to do some really cool things that just aren't practical today, but soon will be.'
Sweeney suggested that there's a pioneering aspect to his work. 'I feel like that's what I'm doing now on Unreal Engine 4 in exploring areas of the technology nobody else is really yet contemplating,' he continued, 'because they're still a few years away from practicality. But I see a huge amount of potential there and so it's very, very fun work.'
When asked about the foreseeable challenges for making the development step up to a new generation, Sweening promptly pointed toward multiple CPUs. 'The big challenge that's going to be coming up in the next decade is scaling up to tons of CPU cores. The way we write software today in Unreal Engine 3 is to have one processor handle all the graphics and it's only a single CPU core with another processor that's dedicated to all gameplay that's running on another CPU core.
'The next challenge is going to be scaling up to tons of CPU cores. But once you have 20 cores, you can't easily say this one is going to be for animation and this one is going to be for details on the face of the character, because all these parameters change dynamically as different things come on screen and load as you shift from scene to scene. So the big challenge will be redesigning our engine and our workload so that we scale more of these different computer tasks between CPU cores seamlessly in real-time and dynamically so that you're always getting the maximum computing power out with the engine, regardless of what sort of work you're doing.'
Sweeney also suggested that, aside from a future milestone of game development being 'movie quality' graphics, human aspects, character simulation and AI would also be key.
'They require increasingly sophisticated algorithms and simulation of human intelligence,' he said. 'I have no idea when those problems will be solved. I'm quite sure they won't be solved in the next ten years. They may not even be solved in my lifetime, but those are all problems that require understanding how the human brain works and trying to simulate that with varying degrees of accuracy. We've seen very, very little progress in these areas over the past few decades so it leaves me very skeptical about our prospects for breakthroughs in the immediate future.'