Anyway, we can conclude that the benchmarks of a ClawHammer processor running at this “ridiculous” frequency were pretty impressive. Note that we mean impressive against the background of other 800MHz CPUs. I suggest taking a look at the table below with the results obtained in Quake3 Arena, which has actually always been a success for Pentium 4 processors (in fact, these are the only results available, so there is not so much to choose from):
| CPU | Core Clock Frequency | Memory | fps |
|---|---|---|---|
| ClawHammer | 800MHz | DDR333 2.5-3-3 | 183 |
| Athlon MP | 800MHz | DDR333 2.5-3-3 | 130 |
| Athlon MP | 800MHz | DDR333 2.5-2-2 | 133 |
| Athlon MP | 800MHz | DDR333 2.0-2-2 | 135 |
| Athlon MP | 1667MHz | DDR333 2.5-3-3 | 210 |
| Pentium 4 | 800MHz | PC800-45 | 111 |
| Pentium 4 | 1600MHz | PC800-45 | 182 |
Look: ClawHammer working at 800MHz shows here the same results as Pentium 4 (Willamette) working at 1.6GHz! Not bad, eh? Bearing in mind that the actual clock frequency of the first mass ClawHammer should be around 2GHz, we dare suppose that its rating (3400+) will be really close to the real state of things (that is to the performance of the “virtual” Pentium 4 with 512KB cache and 3.4GHz core clock). Although we should point out that Intel is not going to have any Pentium 4 (Northwood) processors working at this core frequency. Only 0.09micron Prescott due next summer (see this news story) will be able to reach this working frequency.
And I would like to point out once again that AMD 0.13micron processors are very tough in terms of core frequency increase. You will probably agree that the gap between the today’s 800MHz and the desired 2GHz is anyway too huge, even for a test sample. However, let’s hope for the better...





