Tim Sweeney: We only need what Moore’s law will readily provide. Compared to current-generation consoles, I’d much like to see roughly 8-12x more CPU performance, 10x higher GPU triangle and rasterizer throughput, and 20x more GPU computational (ALU) throughput. With that degree of leap, I’m confident we can ship next-generation games with graphics on par with the Samaritan real-time demo.
g™: What would be the impact on storage media? Are DVD or Blu-ray big enough for the sorts of games that are coming?
TS: Spinning mechanical media is truly an awful distribution scenario for gaming. Accessing a random piece of data takes up to a quarter second, while NAND-based flash memory access times are measured in nanoseconds. Many of the limitations in current-gen games, from Gears of War to GTA, derive from the hardware’s inability to load random-access data quickly. Given that optical drives are also expensive and don’t cost-reduce over time, I’d expect mechanical media to play no role in future entertainment devices.
g™: And how will the march of technology impact the average development studio? Are we talking about a significant increase in expense and expertise?
TS: Budgets of true triple-A titles will continue to march upward. But the game industry has stratified amazingly well in recent years, enabling great games to be developed across two orders of magnitude of budgets. Fifteen-person teams are shipping great Xbox Live Arcade titles, and two-person teams are doing great things on iOS and Android. Web games and social games are doing well with modest budgets. We at Epic expect this trend to continue.
g™: Console generations move in huge steps but PC technology is more of a gradual evolution. With much longer console life cycles now in evidence do you agree that PCs could become the dominant platform in the next few years? And do you think this could lead to more PC-exclusives as a result?
TS: PC gaming tends to make a resurgence toward the end of a console generation, as the performance difference grows larger. This is certainly happening now; witness the rapid growth of Steam-distributed PC games. Whether the trend reverses with the release of new consoles is a function of the quality of the hardware and its appeal to gamers. Certainly, PC is a far better platform now than it was in 2005 when the current consoles staged a takeover of gaming. Capabilities like social networking and video editing, and platform improvements such as SSDs and good integrated graphics even on low-end PCs, mean next-gen consoles will have to truly excel in order to recapture their user base.
g™: The demo you’ve shown is very impressive but it is a real-time cinematic rather than a playable experience. How would it change once other factors are introduced to make it an interactive experience? Can the same level of visual quality be maintained in a real game?
TS: In a cinematic, the artists have complete control over the camera angles, character animation, and fill lights, so those aspects will always be adjustable beyond what happens spontaneously in gameplay. But, the Samaritan demo runs 100 per cent in real-time on consumer hardware you can buy right now. The character detail, particle effects, lighting, shadowing, post-processing capabilities, and overall visual quality are exactly the same when you’re running around the world under player control.
g™: The demo you’ve shown wasn’t just about tech. It had a narrative, character design and world fiction of its own. What was the thought process behind the style and do you have any plans to use the same character/setting again?
TS: Epic’s best and brightest folks spent a lot of time fleshing out the details of the world you saw in the Samaritan demo. As we said at GDC, Samaritan is our proposal to hardware manufacturers. If or when it expands further remains to be seen.
g™: How long did it take to create the demo and what were some of the biggest challenges involved?
TS: Samaritan is the result of three months’ work by a small team of artists and programmers within Epic, as well as NVIDIA engineers who contributed to the advanced DirectX 11 and physics features we demonstrated. This was a pioneering effort, simultaneously figuring out what our development pipeline should be, creating content within that pipeline, and optimizing the visual quality and performance of the end product. We aimed very high, seeking a true movie quality of character lighting (via subsurface scattering and advanced shadowing techniques), reflections, filmic camera effects, cloth, and particle effects. Enabling these features to run with full quality in real time on DirectX 11 hardware required substantial original research by the development team, as well as major code and content optimization efforts.
g™: Games playing devices are arguably more diverse than ever. PS3, 360 and PC are all very similar but you also have Apple and Nintendo devices, plus the NGP. How do you approach development of something like Unreal Engine when not every platform is heading in the same direction?
g™: How much of the new Unreal Engine features is directly dependent on the range of Direct X 11, and its associated power of gaming PCs? What does this mean for its effectiveness on current generation consoles?
TS: Everything we showed in the Samaritan demo is intended to showcase both next-generation console possibilities and DirectX 11 on today’s high-end PCs. Though bits of it – such as cloth and some post-processing effects – are technologically compatible with current-gen consoles, the sheer magnitude of its usage in Samaritan goes far beyond. All of these techniques are available to Unreal Engine 3 licensees and can be readily adopted in PC games today, from MMOs such as Blade and Soul and Tera to PC versions of multi-platform games.
g™: Greater visual fidelity allows you to do much more with characters and facial animation as your presentation demonstrates. Do you think it’s natural that the advance in technology will lead to more emotionally invested, story driven games?
TS: Creating compelling characters that immerse players in interactive environments is one of the biggest challenges for game developers in the next decade. Each generation makes significant advances. Remember the first Unreal’s bots navigating the level just as players do, or the emotional responsiveness of Alyx in Half Life 2? The next generation will need breakthroughs of its own.
g™: Similarly, how do you think the interactive experience will change along with the visuals? Will better graphics allow us to create new types of games?
TS: I’ll go out on a limb and say that better graphics will enable us to create far prettier and more compelling games, but that graphics are no longer a driving force in genre-creation. What genres have been invented in the past few years? Casual social networking games like Farmville; touch-screen object interaction games like Angry Birds and Cut the Rope; new casual microtransaction-based MMOs. Graphics are now sufficient to approximate the visuals required for any genre, and the driving forces are input, connectivity, and creativity.