Well, no, that's not true either, and even if it was, it's sort of moot insofar as the fastest G4 is clocked at 1.33 GHz, whereas the slowest P4 was introduced at 1.4 GHz. In any case, to revisit the record, the post I was responding to was #43, which claimed "a G4 Mac is typically about double the speed of a Pentium IV with the same clock speed." Sorry, no.
Quit twisting words. By "same clock" I mean "clock-for-clock," same thing if you get two at the same clock. And you can. The fastest MPC 7xxx ("G4") that Apple sells is 1.5GHz (not 1.33), and it will absolutely blow away a P4 1.5GHz. Since the original P4 was slower than even a lower-clocked PIII, this latest G4 is probably more than twice as fast as that quite old Pentium chip. But that's not a comparison of what's on the market now, so I'd consider that invalid.
Still, clock-for-clock, a G5 or G4 is faster than a Pentium, same as with AMD chips (I remember an Athlon 1100 trouncing a P4 1.5 in tests). The real reason for this is that when Intel made the P4 they counted solely on high clock speed for performance, for a long time not bothering to improve the microarchitecture to make it more efficient for each clock. This worked well for them for the couple years following the P4's release, but the speed curve got pretty shallow around 3GHz.
Following recent news, including cancellation of the 4GHz P4, Intel has pretty much admitted that this strategy has hit a wall. IBM and AMD (who have a fab technology sharing agreement) already have more efficient designs for their chips, and can slowly bump clock speeds to stay ahead of Intel while they play catch-up.
which claimed "a G4 Mac is typically about double the speed of a Pentium IV with the same clock speed." Sorry, no.
I don't agree with that either. Maybe on very specialized SIMD operations, but otherwise, no.