At 02:08 PM 4/12/2002 -0400, Dan Sugalski wrote: >At 12:49 PM +0200 4/12/02, Paolo Molaro wrote: >>On 04/11/02 Dan Sugalski wrote: >>> I'm not sure which is worse--the amount of data we're copying around, >>> or the fact that we eat Mono's lunch while we do so. >> >>:-) >>Could you post the code for the sample? > >Simon did a one-for-one translation of life.pasm to C# and we were rather >significantly faster. (I'd much rather use .NET numbers, but we don't >currently have them) > >FWIW, the numbers were: > >No JIT: Parrot 866 gen/sec Mono 11 gen/sec > JIT: Parrot 1068 gen/sec Mono 114 gen/sec > >We can't find the original--I've attached the version he's got now, but >it's apparently even slower.
Also don't forget the version I wrote for cola, its similar to the C# one. I just realized Cola is broken in the current CVS version, so I'm flushing my fixes/changes before I leave. life.cola is working again and is actually faster than life.pasm %) Try timing it. Likely it is because it uses the replacement version of substr op for string indexing. For fun, you might take the a.pasm generated by Cola and patch in the timing and interpinfo ops to see how GC compares for my version. I'd expect it to be much lighter. -Melvin
