GG: Will there be the facility to Scan Line Interleave.

TW: No there won’t. We don’t believe in that…the reason single chip implementation has more than enough rendering performance for Intel CPU’s way into the middle of next year and putting two chips into the same system will literally give you no benefit. With the second generation chips we can support 1,600 x 1,200. SLI is a fill rate issue and we have more than enough fill rate.

GG: How are you going to overcome the dominance of 3dfx, after all they'll already have huge numbers of Voodoo2s in the market by the summer.

TW: The reality of the situation can only be described by the numbers of chips shipped and up until the third quarter of last year we had shipped more chips than 3dfx had shipped, that’s PowerVR versus Voodoo 1. In the last quarter of the year they shipped a huge amount more than they shipped for the three quarters and the latest data shows that Voodoo is slightly ahead of us but not by very much with a 29 per cent market share and we have 24 per cent. Voodoo and PowerVR are the major preferred chips by volume of shipment into the market.

The second issue is the market perception and the market perception is higher for 3Dfx. Also developers are moving away from proprietory APIs to DirectX6 which is the favoured API, which is their favourite API for the coming year. Our feeling as we move forward is that games developer issue is going to be less of an issue…the deciding factor will be the price points of the chipsets…if you want to get into the volume mainstream business of the PC you need to have a chipset that’s in the $30 range.

GG: So how much will the 2D/3D combination sell for in Britain?

TW: About the same…around $30 for the chip and that’s where PowerVR really picks up…it’s price/performance is impossible to match because the architecture is different from almost every other type of 3D chip.

GG: What’s been the reaction of companies such as Gateway, Dell, etc.. to incorporating cards based on PowerVRSG into their machines?

TW: Response has been good because it’s in the price arena that they want…it provides high end functionality and high end performance.

Click to continue . . .

GG: So who’s going to support it then?

TW: We can’t tell you that for obvious reasons…when they announce then we will announce at the same time.

GG: When are we going to see boards in the stores?

TW: Around August time frame.

GG: Moving on to the future of graphics technology, how do you see the industry developing over the next 10 years?

TW: First, the echnology is not the issue here. Whatever is available for the workstations is also available for the PC…but at a price point but a price point too high for it to transcend into the PC space right now.

Secondly, the workstations have much higher floating point performances than PC processors, Pentium processors have. That’s why workstations are where they are.

Intel’s plan for their next generation of chips are to deliver basically work station chips. When that happens the distinction between PCs and workstations will be very grey. In a couple of years time you are going to see PCs that can deliver more than what workstations can deliver today. Workstations will have to move up the ladder to have a space to play in and so they (future PCs) are going to get to the point where they really are going to do Jurassic Park stuff in real time.

GG: So can you predict what will be happening in 2005?

TW: Workstations will be dramatically more powerful than they are today…there will be less of them, fewer companies competing it’ll be more specialised and I think that high-end PCs will undoubtedly have taken over from the low-end workstations.