Competition

R&D and the competition

By now everyone knows the release schedule that NVIDIA has enforced upon themselves. Releasing a new product every six months is pattern that NVIDIA has been keeping up with very well but unfortunately, most consumers can’t say the same for themselves. It’s definitely exciting to see new things come frequently, but many people are also becoming increasingly frustrated because their investment becomes outdated in a matter of weeks.

A good reason why the graphics industry should slow down is to allow games to catch up and fully utilize the graphics processor. Now, you may think that if this happens, games will become too complex and the cards won’t be able to handle them. But on the contrary, the exact opposite may happen.

Take for example the PlayStation2. It’s a closed architecture system, meaning you can’t upgrade its critical parts like you do with a computer. Yet, seemingly the games just keep getting better year after year and the PS2 will remain a viable console for at least two more years. This happens because developers are forced to design their games within boundaries set by hardware. They must conform and flex their skills harder when developing for a console than for an open platform like the PC, and over time they learn more and more about how to best utilize the hardware.

A year, or more

With Parhelia, Matrox has told us that it plans to have it be a successful product for at least a year. This probably means that Matrox plans to have a design and produce period of at least a year for new products. As long as each offering is powerful enough to last the entire cycle, it shouldn’t be a problem for Matrox.

Matrox’s strategy actually isn’t too far off from NVIDIA’s and ATI’s. What NVIDIA and ATI do is produce a new architecture every year, with refreshes half way. Typically we will see a new product every spring season and product refreshes in the fall. This is when a product will be updated with, say, faster memory or a faster core clock frequency. The internal architecture itself though, will remain largely the same, if not exactly.

Going this way, there won’t be a need for mid-year refreshes and customers can hold on to their investment longer and won’t feel like they’re being ripped from their money. However, the only way to get the true value of a graphics card is to pair it up with applications that utilize it. With at least a one-year life cycle, the Parhelia will see its feature set being increasingly exposed to applications and games, which means the end user will be continually rewarded. Before T&L was even being considered for support in games, NVIDIA’s GeForce256 was already out the door.