Watch.Impress have posted their interview with Nvidia's chief scientist David B. Kirk in two parts... Here's a quick translation as the site's in Japanese:
Q: The NV30 (the GeForce FX) the clock, 500MHz is very high the NV25 (the GeForce 4 Ti) with by comparison. It is due to refining the process, with just shortening the gate length of the transistor in regard to calculation, is the expectation which does not reach to the 500MHz. With the GeForce FX improving pipeline architecture destined for acceleration?
A: Pipeline change was done? Well, it has not done. Speed has improved purely with just process technology. The largest reason is copper wiring. As for 0.13 mu m process of the TSMC copper wiring is used. With that, it could designate high clock conversion as because wiring resistance goes down largely.
GeForce FX Secret Of High Performance
Characteristic And Efficiency Of The GeForce FX
Q: The NV30 (the GeForce FX) the clock, 500MHz is very high the NV25 (the GeForce 4 Ti) with by comparison. It is due to refining the process, with just shortening the gate length of the transistor in regard to calculation, is the expectation which does not reach to the 500MHz. With the GeForce FX improving pipeline architecture destined for acceleration?
A: Pipeline change was done? Well, it has not done. Speed has improved purely with just process technology. The largest reason is copper wiring. As for 0.13 mu m process of the TSMC copper wiring is used. With that, it could designate high clock conversion as because wiring resistance goes down largely.
GeForce FX Secret Of High Performance
Characteristic And Efficiency Of The GeForce FX
