Q&A: Ageia's Steele On The Battle For In-Game Physics

How is Ageia's plan to revolutionize the game industry with a standalone physics hardware card going? We caught up with VP of marketing Michael Steele to discuss the pivotal year Ageia has ahead of it, and how the company can win over developers and consumers alike. The company has been continuing to forge towards its vision of mass-market adoption, most recently announcing both expansion of the company's processing units into mobile hardware, as well as working on original games that take advantage of Ageia's specialized physics engine, including the action title CellFactor and Warmonger. In the course of our conversation, originally recorded for a GDC Podcast by Tom Kim, we discuss the competition Ageia is seeing from ATI's Crossfire and Nvidia's SLI boards, and how the company can make the case to core gamers that its hardware can provide solutions for real, perceivable improvements in game mechanics, beyond visual effects. Gamasutra: Ageia is best known for its PhysX hardware solution for gameplay physics. So you've got this piece of hardware that costs roughly $200. What is going to justify spending that kind of money? What's going to make it an indespensable peripheral that gamers are going to want to add to their computer? Michael Steele: The holy grail of physics in games is game-changing physics that impact gameplay. Not simply effects, but gameplay-changing. What we're seeing is a steady wave of content that's changing from the point at which we launched through today, on into 2007 with game-changing elements in upcoming games like CellFactor, Warmonger, Ubisoft's Ghost Recon 2, and a number of games built on the UE3 engine from Epic, which we're integrated deeply into. So the justification for providing a physics processor is that kind of game-changing environment. In the last year, we've become very focused on making that happen. What we're seeing is a steady stream of content that will roll out through the rest of the year that will make it both a compelling hardware and software solution. Gamasutra: A lot of people are licensing the UE3 engine, but your solution is as much pull as it is push. The developers have to actually make use of gameplay-changing physics in order to really show off what your hardware solution can do that other people can't do. What are your efforts in terms of addressing the development community to say, "Hey, maybe you should look at this and make software for this that will broaden the experience for gamers?" MS: You're right, it is a pull and push situation. Specifically to UE3, I think the best way to articulate it is to look at a specific example. We've worked very closely with some technical minds at NetDevil, the developers of Warmonger. It's the first UE3-based game that is specifically designed with exclusive use of the PhysX engine, and specifically designed to take extreme advantage of the PhysX processor. So you're going to see things like ember effects, which are floating around the environment at all times. You're going to see what we call "metal cloth," wrapped around vehicles, so when you hammer the "cloth," it actually reacts as you would expect the side of a car to. You're going to see a totally destructible system, which means that every piece of the game itself is totally destructible. You can blow holes in the sides of buildings; you can blow holes in things in front of your path to create new pathways for you. It actually puts more intense pressure not only on the rendering requirements of the graphics processor, but also the AI requirements of the CPU. It takes full advantage of what we call the "gaming power triangle," which is what we call a complete gaming platform composed of a CPU processor, a graphics processor, and a physics processor. It does a couple of things. It gives gamers a great game to play that takes full advantage of the physics processor, but it also provides an excellent platform for developers to lean back on as they move forward in developing more creative implementations of PhysX games. Gamasutra: Part of that triangle is the graphics processor -- standalone hardware that does graphics processing. But you've got some pretty stiff competition out there. The Havok solution, for example, makes use of multicore and multithreaded processors on GPUs from both NVIDIA and ATI. Also, processors on motherboards themselves are now multicore and multithreaded, so developers have other choices for physics implementation. It takes some central load off of gameplay effect to give you some of those same effects. So, why should they choose AGEIA's solution versus some of the other ones that are already there, particularly the ones with GPUs that people are already going to buy because it's become more or less of an indespensable high-end addition? MS: As a developer, you have three choices for developing physics. You can do it on your own -- and there's some guys out there who do a pretty good job of it -- or you can buy it from someone like Havok for $100,000 or more. That's an expense that might be worthwhile, because I'll admit, I have it because it's a nice software solution. It's not built for hardware, though. It's not built for PhysX-optimized hardware. From a graphics standpoint, what you need to think about is what games today are running on Havok software on a graphics processor. Take a look at that list, and that will give you a sense of what the hardware opportunity is on the graphics side. I think what we're all about is providing a complete solution from top-to-bottom on the software side and the hardware side. We provide a compelling hardware architecture that takes physics far beyond what you can do in any existing architecture that is built for graphics or general purpose processing. We provide a complete solution, which is rounded out by our software and middleware. We spend a lot of effort, and a lot of time and money on the software side, and what results from that is a royalty-free SDK. From a developer's perspective, paying nothing for the software is somewhat compelling, and having the opportunity to take that software and optimize it for purpose-built hardware gives you the opportunity to set yourself apart from the rest of the pack with game-changing physics. Gamasutra: You've mentioned to me that that royalty-free SDK is available on all the new generation game consoles as well. However, those are relatively fixed consoles, and might have limited options as far as integrating a hardware-based solution goes. You mentioned that maybe something can be done in that area as well. MS: It makes sense as a provider of physics software that we develop a cross-platform solution. As a developer, you want to take a solid piece of software and be able to apply it to as many platforms as possible. While it's true that we really make our money on the PC platform because we make hardware for the PC, we also provide a mechanism by which developers can build once and cross-compile multiple times, across the PS3, Wii, Xbox 360, and PC. We have to provide a complete top-to-bottom solution, and that's a solution that means something for the gamers, but it's also a solution that is useable and something that's fair to implement by the developers themselves. Within the console space itself, there are games on every console today that utilize the PhysX software engine. We don't provide the hardware for that, but from the software side, we provide a compelling top-to-bottom solution. Gamasutra: You have a specialized solution which is optimized toward provision of physics. To some degree, though, gamers still have to care about it. In other words, there has to be some kind of draw, like a triple-A title where PhysX is an essential part of the product. Development chases the audience, and the audience has to be somewhat compelled by your product. Of course you've got integration with UE3, but can you talk about any titles that are currently in development that might be system-sellers for you? MS: One of the things we have to be careful about is that we can't pre-announce titles of our key partners, but I can talk about three things that are very interesting right now. CellFactor was the most popular game demo when we launched our product, and now it's being turned into a five-level minigame which will take full advantage of the PhysX processor. That will be rolling out very shortly. On top of that, there's also Warmonger from the guys at NetDevil. That will also be coming out in a matter of weeks. It will be four to five levels of a minigame which is taking full advantage of the hardware. It will be an example of the first implementation of a UE3 engine-based game. Of course, that leads right into UE3, not the least of which is Unreal Tournament 3, a game which we expect to be very highly anticipated. From a physics perspective, there is a lot that we can add to an already exciting franchise. But beyond that, we're also working closely with the guys as Ubisoft on Ghost Recon 2, and we can expect to see some exciting stuff from them in the spring or summer as that game rolls out. We've fundamentally changed the way we approach development with that game, and we've made significant changes to not only the underlying PhysX SDK -- given that the Ageia PhysX engine will be the exclusive engine on Ghost Recon 2 -- but also we are taking full advantage of what we can do there. Gamasutra: One criticism regarding the PhysX implementation in the original Ghost Recon was that it really was more of an effects-based physics rather than gameplay-based. It's important that developers start implementing more gameplay elements that have to do with physics, rather than just the visual impact. What kind of work are you doing with developers to help them understand how to make better physics-based gameplay? MS: There are a number of things. We provide both the hardware and software, obviously. On the software side, we've continued to make new updates and revisions to our SDK. I think most developers, if you've asked them today, our SDK as compared to where it was 12-18 months ago is much different in terms of usability and the tools we provide. The other way you need to look at it is not only are we on a curve of improving and maturing our software, but the developers themselves are also learning to use our physics as well as learning what is the most creative way to bring PhysX into a game. It's a new shift in how you look at bringing game-changing elements into your design, and it takes a while to figure that out. We're at the point now where we've figured out quite a bit from the software side. We've already got a strong piece of hardware there waiting we could utilize more fully, and the developers themselves are learning more as well. The other thing I'll say is we've become somewhat more focused on who we are working with and how we are working with them. We've been putting more resources into the game development itself from both a financial and personnel perspective. We're making it happen from our end, and we're working very closely with those guys we think can make the most impactful games out the chute. I think from that, what we'll find is games that become more impactful and more exciting, and will drive additional developers to see what they can do when they see the potential. That leads to a groundswell of support in the future. Gamasutra: We're looking forward to seeing what developers can do with PhysX, and we're curious as to how this will all play out. Thank you very much for your time. MS: Thanks! We're looking forward to an exciting 2007 and beyond.