Some weird benchmarking numbers

Dmitry Borisov jbors at mail.ru
Wed Aug 18 23:48:31 CEST 2004


Guys,
Just get back to the shootout site and was redirected here:
http://shootout.alioth.debian.org/index.php
After I saw psyco having the same benchmarks as Python I tested that by myself and found those numbers to be totally odd.
Here are the tests I've made so far:

( http://shootout.alioth.debian.org/bench/lists/ )
c:\bors>list.py 100
10000
Without psyco  5.00699996948
10000
With psyco  3.02399992943

( http://shootout.alioth.debian.org/bench/sumcol/ )
c:\bors>sum.py 100000
4999950000
Without psyco  0.0999999046326
4999950000
With psyco  0.0599999427795

( http://shootout.alioth.debian.org/bench/ary/ )
c:\bors>ary3.py 10000
1000 10000000
Without psyco  5.66899991035
1000 10000000
With psyco  1.1210000515

I'll try to contact those guys to have it corrected. Anyone volunteer to have other tests compiled ?
Thanks,
Dmitry/



More information about the Python-list mailing list