This is an excellent news and the benchmark is a realistic action! Thanks you all! On Jul 12, 2010, at 9:01 PM, John McIntosh wrote:
As always it's difficult to determine how best to benchmark. So given:
| savedMorphs | savedMorphs := World submorphs. World removeAllMorphs. self halt. ^ [[10 timesRepeat: [Browser fullOnClass: Object selector: #yourself]. World submorphs do: [:m | m delete. self currentWorld doOneCycle]] ensure: [World addAllMorphs: savedMorphs]] timeToRun inspect
I can say that the Cog Stack VM on the iPhone produces numbers like
48.081 seconds 47.858 seconds
The original VM did:
74.415 seconds 74.257 seconds
making the iPhone COG VM 1.55x faster
A 2.33ghz mac book pro mac vm 5.7b1 does it in about 8.3 seconds with a COG JIT VM 5.8 that takes 2.6 seconds or about 3.2 faster
-- =========================================================================== John M. McIntosh <johnmci@smalltalkconsulting.com> Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ===========================================================================
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project