While running a quick benchmark in Quake3, 1600x1200, 4XAA and 8xAF it got 111 FPS, compared to a GeForce FX Ti5800 Ultra that got 48 FPS. However both chips were clocked at 250MHz because it was still a prototype, and they wanted to make an exact comparision to the NV30. So basically both were set to 250 in order to make a fair benchmark.
I can't assure how accurate this translation is (see original article), perhaps some of our german visitors can give us a better hint. From what I could also understand, the NV35 didn't need any special cooling method despite of the high frequencies it was working at, though it could have features disabled as it was a beta board. Also the performance gain is expected to come from the wider memory bus, up to 256bit (from FX's 128), and further core optimizations with some 5 million transistors added to the mix.