Oddity about declared multi-core CPU power?

Question

I am having quite a hard time in understanding the output of cat /proc/cpuinfo on two 4-core Linux boxes here at work. Basically I am running algorithm experiments on the two machines (let's name them red and blue); on the average, when running the same algorithm, red takes more (CPU) time than blue to get the same results. The outputs of cat /proc/cpuinfo on the two machines is (for all four cores):

Blue

processor   : 0
vendor_id   : GenuineIntel
cpu family  : 6
model       : 23
model name  : Intel(R) Core(TM)2 Extreme CPU X9650  @ 3.00GHz
stepping    : 6
cpu MHz     : 1998.000
cache size  : 6144 KB

Red

processor   : 3
vendor_id   : GenuineIntel
cpu family  : 6
model       : 15
model name  : Intel(R) Core(TM)2 Quad CPU           @ 2.40GHz
stepping    : 7
cpu MHz     : 2403.000
cache size  : 4096 KB

So, the results seem somehow related to the frequency as given in the model name row, while the value in cpu MHz seems to counter this trend. How should I interpret these data? How is it possible that the single core power of blue is lower than the single core power of red and yet the overall power of blue is higher? Could someone redirect me to an article explaining the phenomenon?

Thank you
Tunnuz

Answer 1

You should heed the frequency given in the "Model name" field. At the time you took this snapshot of the CPU info, Blue happened to be running with a lower multiplier, likely because it wasn't under any load.