The good doctor Con Kolivas [interview] recently posted some interesting kernel benchmarks generated with his contest benchmarking tool [story]. Using resources provided by the OSDL, Con compared 2.4.19, 2.4.20, 2.5.49 and 2.5.50, each with a single processor enabled and with dual processors enabled.
Below many of the tests, Con makes some quick observations about the results. Con’s original email is followed by an edited version I’ve added to hopefully make it easier to compare the results between UP and SMP. Read on for the hard data…
The details are in the rest of the article, but as far as I can see there are some things that 2.4 does slightly faster but other things that 2.5 does significantly faster. I’m sure hiccups will be ironed out by the time 2.6/3.0 is released.