Re: [DotGNU]CVM to machine code

[Paolo Molaro]

On 07/27/03 Rhys Weatherley wrote:
> On Sunday 27 July 2003 04:19 am, Paolo Molaro wrote:
> > Mono is about 3 times faster than the fastest ilrun build, the others
> > are 5 or more times slower.
> 
> I am curious - how much faster is Mono than Mint using similar tests?  Also 
> Rotor?  And Microsoft's commercial engine?  Right now, there are only two 

As you know, you can't distribute scores for the MS engine without their
approval and I'm not going to bother asking for it. Try it out on your
windows PC. As for Rotor, I don't have or use it, so maybe you can run
the tests with it (dunno if their license allows it, either). Also, if
you want to spend the time to benchmark mint, just do it, I consider it
an exercise in futility:-)
That said, someone published some benchmark results for scimark and
other stuff, comparing MS 1.1, rotor and an old version of mono.
Current mono should be about 50% faster than the numbers they give.
See
http://weblogs.cs.cornell.edu/AllThingsDistributed/archives/000078.html.
In scimark ilrun is about 9 or 10 times slower than mono, so, if you
extrapolate from the graphs, ilrun should be slightly faster than rotor
(but apparently the rotor people contested those results, saying they
weren't done with a release build of it).

lupus

-- 
-----------------------------------------------------------------
address@hidden                                     debian/rules
address@hidden                             Monkeys do it better