NVIDIA Introduces GeForce FX (NV30)

Kudos to ATI, we didn't think that they could pull it off but this year will come to a close with the Radeon 9700 Pro being the fastest card out on the market. We honestly expected, just like NVIDIA, to see NV30 on store shelves (albeit in limited quantities) before the end of December but as fate would have it that's not how things worked out.

Why isn't NV30 here now and why won't it be until February 2003? We've discussed this time and time again and it directly relates to NVIDIA's decision to make NV30 a 0.13-micron design from the start, instead of introducing a 0.15-micron part initially and refreshing it 3 - 6 months later with a 0.13-micron die shrink. Whenever you pursue a not-so-mature manufacturing process (TSMC barely shipped any 0.13-micron wafers in 1H02) over a more mature solution, there are bound to be issues getting yields up to par. Unfortunately for NVIDIA those initial problems have forced NV30, now officially known as GeForce FX to slip into 2003.

With that out of the way, we're finally able to tell you everything there is to know about GeForce FX. To tell the truth, we've been sitting on this information since March of this year and very little (if any) has changed in the specification. NVIDIA had the design and the features of the GPU ready very early this year indicating that it truly was manufacturing that held them back.

The GeForce FX won't be out until almost a full year since we were first presented the specifications of the GPU but just like the past several GPU launches we've seen, it is good to get an initial understanding of the technology before bringing performance numbers to the public. This way expectations are more realistic and you'll know almost exactly what, in your mind, may or may not be worth waiting for come February.

So in the AnandTech tradition, let's dissect NV30…