You were right with largeExplorers benchmark. Tested with 3.8.6665-full.
1) Intel Pentium 4 Mobile - 1.8GHz (WIn)
arithmaticLoopBenchmark 2179 compiled 158 ratio: 13.791 bytecodeBenchmark 2951 compiled 622 ratio: 4.744 sendBenchmark 3402 compiled 2281 ratio: 1.491 doLoopsBenchmark 1920 compiled 3660 ratio: 0.525 largeExplorers 1913 compiled 2013 ratio: 0.950 compilerBenchmark 1855 compiled 1448 ratio: 1.281 Cumulative Time 10830.891 compiled 4739.508 ratio 2.285
2) AMD Semprom 3100+ 1.8GHz (Win)
arithmaticLoopBenchmark 1647 compiled 117 ratio: 14.077 bytecodeBenchmark 2746 compiled 635 ratio: 4.324 sendBenchmark 2097 compiled 892 ratio: 2.351 doLoopsBenchmark 1265 compiled 895 ratio: 1.413 largeExplorers 996 compiled 721 ratio: 1.381 compilerBenchmark 991 compiled 1000 ratio: 0.991 Cumulative Time 6526.543 compiled 2119.418 ratio 3.079
Man, I was born for statistics ...:)
Jakub
On 11/16/06, bryce@kampjes.demon.co.uk bryce@kampjes.demon.co.uk wrote:
One more question. How safe is it to stick with the exupery vm for my non-exupery-loaded images?
Loading Exupery from SqueakMap is perfectly safe. The main reason I provide the pre-build VMs is to make it easier to get started. They are also a useful basis for macro benchmarks as the code behind the benchmark will be the same.
I suspect that some of the differences with largeExplorer performance was due to image differences. I benchmarked using 3.8, Andy used 3.8.1, and Jakub used the pre-built 3.9.
Bryce _______________________________________________ Exupery mailing list Exupery@lists.squeakfoundation.org http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/exupery