The real reason that the Pentium 4 underperforms!

[Novdid]

The original Pentium would need a much higher bus than the original 50-60MHz to be able to perform the same clock for clock at 2,2GHzas it did running at 90MHz. Lets not forget memory prefetching and a whole lot more tweaks it would need to perform adequately.

[Ribbit]

Yeah, that's why I said "other things being equal". I assumed that if we had a Pentium running at 3GHz it wouldn't have 70ns EDO memory on it

[Kooldino]

Yup. Like I said, "with a few minor tweaks"...those being some of them.

[Indiana]

The 68060 was very strong, considering the low clocking, outperforming the first PowerPC 601 in some cases.
Of course the performance vs. the Pentium was dependent of the used application. In the ByteMark benchmark, it got an integer score of only about 0.6 (this score is relative to a P90), so it would've been about the same speed clock for clock. Can' seem to find some SPEC scores for comparision, but I'm quite sure those would favor the Pentium (SPEC scores always favoured Intel, but then this up to now has NEVER translated into an equivalent performance delta in real-life apps, so those SPEC scores are IMO quite meaningless)

In some real apps based on the same source code (e.g. LAME, RC5-64, POVRAY...) the 060 got speeds comparable to a P90.

Those statements are of course also highly machine-specific, since there are quite different memory controllers used which affect performancequite a bit.