For the sake of comparison I have listed all of my benchmarks for the 188 pointers below.
Project : 2306
Core : GBGromacs
Frames : 100
Credit : 188
2.4 P4 @ 2463MHz
Min. Time / Frame : 26mn 21s - 102.74 ppd
Avg. Time / Frame : 28mn 04s - 96.46 ppd
1.6 P4M Laptop
Min. Time / Frame : 24mn 25s - 110.88 ppd
Avg. Time / Frame : 25mn 25s - 106.51 ppd
A64 3500+ @ 2376MHz
Min. Time / Frame : 15mn 05s - 179.48 ppd
Avg. Time / Frame : 15mn 35s - 173.72 ppd
A64 4000+ @2592 MHz
Min. Time / Frame : 15mn 15s - 177.52 ppd
Avg. Time / Frame : 15mn 15s - 177.52 ppd
A64 X2 4400+ @ 2310 MHz (2 clients - other one has a 46 pointer

)
No Min. Time / Frame
Avg. Time / Frame : 14mn 22s - 188.44 ppd
As you can see the A64s take nearly half the time the P4s do but nowhere near 6 days.
MajorPart:
As Rich says, you may need to check that the optimisation is kicking in. Are you using the -forceasm flag?
Stan
