I'm trying to figure out why the 800 MHz Celeron II's in my fleet out-perform my 2.0 GHz and 2.4 GHz Xeons. 
I'm running the Truxoft clients on all of my systems. The Xeons are using the clients optimized for SSE2 and the benchmarks were run with all other processes stopped.
Here are benchmark scores for my IBM eSeries 2.0 GHz Xeon dually:
1127 double precision MIPS (Whetstone) per CPU
1710 integer MIPS (Dhrystone) per CPU
Here are benchmark scores for my Dell 2650 PowerEdge 2.4 GHz Xeon dually:
1398 double precision MIPS (Whetstone) per CPU
1933 integer MIPS (Dhrystone) per CPU
Now, here are the latest benchmark scores for one of my 800 MHz Celery II's:
746 double precision MIPS (Whetstone) per CPU
2035 integer MIPS (Dhrystone) per CPU
Is this normal? Granted, the Whetstone score is considerably better for the Xeon but I would have expected the Xeons to run circles around the Celeries in both tests.

I'm running the Truxoft clients on all of my systems. The Xeons are using the clients optimized for SSE2 and the benchmarks were run with all other processes stopped.
Here are benchmark scores for my IBM eSeries 2.0 GHz Xeon dually:
1127 double precision MIPS (Whetstone) per CPU
1710 integer MIPS (Dhrystone) per CPU
Here are benchmark scores for my Dell 2650 PowerEdge 2.4 GHz Xeon dually:
1398 double precision MIPS (Whetstone) per CPU
1933 integer MIPS (Dhrystone) per CPU
Now, here are the latest benchmark scores for one of my 800 MHz Celery II's:
746 double precision MIPS (Whetstone) per CPU
2035 integer MIPS (Dhrystone) per CPU
Is this normal? Granted, the Whetstone score is considerably better for the Xeon but I would have expected the Xeons to run circles around the Celeries in both tests.