Sony's Phil Harrison on PlayStation 3

Everything in the demos was real-time.

Not all of that - in fact, none of it was real-time because it was all running off video. If you make a presentation to two and a half thousand people, you're going to put some of it on video just to be on the safe side.

I've been asked this question a lot. The way we put those videos together, everything was done to specification. Everything was done to PS3 spec. Virtually everything used in-game assets; some things were rendered.

I think very. I think depending on the game, different games took a different approach to their way of expressing what the games are like - but clearly, something like MotorStorm uses more cinematic, replay-like cameras than you would ever enjoy in-game. So that makes a big difference... But everything is done to spec.

I just think we had great stuff to show! Yeah, I'm really proud of the way the European content has been received, and I'm delighted with the response to Heavenly Sword, MotorStorm and Killzone in particular.

But even things like The Getaway technical test - and I was at pains to point out that this was not Getaway PS3, this was what happens when a team rolls off a game and we start getting them thinking about what is their vision for taking that technology and scaling it up. I think it was a good A to B comparison, because people know what Getaway looks like on PS2, and then they were able to ramp that up on to PS3 - albeit on very early prototype hardware, so it was a bit painful along the way for them! I thought that was a great example.

Yeah, good example - I mean, you could see actually, the way that those cameras worked. That zoom-in camera was done in real-time to capture the kind of video-like footage that we had.

It's really hard to say, because as technology gets more and more complicated, there's no concept of the "perfect" engine. We used to say on 16-bit that a game used 90 per cent of the machine's power, or Gran Turismo uses, you know, 94 per cent of PlayStation 1's power... There's no concept of the perfect game engine that uses everything. So it's hard to say.

It varies between developers, because we've got different variants of the hardware with different performance characteristics. Obviously as you get closer and closer to production hardware - and typically, the final devkit that you get is production hardware, near as damn it finished - as you get closer and closer, you're using more and more like the final silicon, which will be more and more like the final clock speed.

So it only gets better from here on in - which is pretty astonishing, to think about the implications of that actually. But faster, more powerful - where you use that power is a different issue.

You're right; obviously Cell allows you to do complex collisions, physics, dynamics, simulations, all of those things. Though, the Getaway demo was a good example of how you can have a living city brought to life as a result. Although it was pretty graphics, most of that power was actually Cell-based.

The Doc Ock head - the Alfred Molina head - is actually more of a Cell demo than it is a graphics demo, because we're calculating hugely complicated light sources in real-time on the Cell, even to the point where we calculate the angle at which light enters the skin, the way that the light is then coloured by your blood, and the way that it is then reflected back out. It's something called transmission. Skin is hugely complicated - if I put my finger over a light, for example, you can see that the light is coming through my skin. We were simulating that - emulating, simulating, kind of a fine line - we were simulating that on the Doc Ock head demo.

Yeah. Those are really hardcore maths problems which the Cell is really good at solving.

Well, I'll give you a couple of other examples. The terrain-rendering demo that was done by STI, which is the people who developed the Cell, doesn't use the graphics chip at all. That 3D landscape was generated in real-time from two input data sources and a software renderer running on the Cell created the final image. All that it does is output as a bitmap straight to the video hardware - it doesn't even create a single polygon, there's no concept of a polygon in that demo.

It varies from two to probably five months. We've had Cell for a bit longer than we've had the graphics chip - at least, a working graphics chip. We've had our devkits for just over five months now.

It always made me chuckle, that comment from Microsoft, because yeah, it's true, but it didn't stop us having thousands of games and 80 per cent market share. Having said that, there is an element of truth to it - PlayStation 2's architecture was more challenging for your average developer to get their heads around. Some were capable of getting their heads around it, some weren't.

PlayStation 3, I think, is going to be cheaper to develop for than the corresponding period of PS2 development. I know that's a fairly contentious statement to make, but there's a very good reason for that. When we announced the collaboration with NVIDIA, we just talked about them making a chip - actually, they don't make anything, they're a designer, and the RSX contains an NVIDIA-designed part, which gives us fantastic GPU capabilities. But what it also gives us, and this is actually the most important bit of it, is all the toolchain and CG pipeline that comes with it, which is a very well understood development pipeline in the PC community - and, yes, in the Xbox community, frankly.

So all of that pipeline of tools and technology and plug-ins comes straight across to PlayStation 3. Plus, on the Cell side of things, IBM brings a lot of expertise and know-how to the table. Also, as you know, the PS2's EE had two microprogrammable devices, VU0 and VU1 - which were incredibly fast, incredibly powerful chips that were very difficult to program for because of their very specialist nature and the programming skills required.

Within PlayStation 3, the Cell chip, although it has a number of components inside it, they're all general-purpose CPUs. They can be programmed at a much much higher level.

Absolutely. Messing around with VU code... Yeah, it's true. It's not for the faint-hearted, for sure.

Well, clearly Monday and Tuesday have been our big coming-out parties. We're now public, so we can now be a lot more open with all of our partners about what we're doing. You'll see a lot more devkits rolling out - but exact schedules, who they're going to and what they do is not something I can discuss here.

Yeah, for sure. We've been making them for some time, but obviously they're not in abundant supply at the moment.

Because it can be a hub, rather than just being a terminal at the end of a network. Also, we want to be able to have a Gigabit port for an IP camera. So one of the ports is an 'in', and two of them are 'through'. It can be a server as well as a terminal.

Depending on the features that you exploit, some of it's handled by the OS, some of it will be handled by the applications. I should also explain that although yes, there are two HDMI outputs, you don't have to have only high-def devices attached to PlayStation 3 - there's also a standard PlayStation AV Multi-Out connector. So one of them could be an HD output, and one of them could be an AV Multi going to the TV.

The Cell can run multiple operating systems, so yes, you could do that. Now, we don't have the application up and running yet, and the resource management isn't quite final - but the purpose of that presentation was to show what is possible. Exactly how that gets unlocked is still being worked on.

But more people play online games on a PlayStation 2 than on any other game console.

Yeah, but it is something that is worth pointing out - although, personally I have a great deal of respect for what Microsoft has done with Live, and I think they've got a lot of it right.

I think that philosophically, PlayStation 2's online offering is an add-on to the hardware and software and operating system. In PlayStation 3, online is part of the DNA of the machine - in fact, the Cell processor itself is designed from the ground up to be connected to a broadband network.

So from switch on, day one of the machine, network functionality will pervade every aspect of the machine. We talked a little bit about that at the conference, but there's a lot more detail to go into about some of the really cool things - like, if your PS3 is switched on at home, it can be a media server to your PSP on the other side of the planet. Now that is incredibly cool. I could sit here in LA and navigate the data which is stored on my PS3, and download music and other data off my PS3, onto my PSP.

Thank you.