[Mono-list] JVM performance: JVM as a basis for CLR

[Jens Bäckman]

Tom wrote:

>On the "fib" benchmark (mostly function calls), CLR
>performs considerably worse than Sun's JVM.  On the
>convolution benchmark, they perform about the same.
>
I tested this code on my own machine with IBM:s JVM, which has included 
a very efficient JIT from version 1.1.6 first released for OS/2 Warp. 
Compiler parameters were identical to yours, machine is a 1GHz Athlon 
with Win2k.

===== Log begin =====
D:\Temp\Performancetest>java fibj
starting
fib(40)=165580141 took 2.584
fib(40)=165580141 took 2.604

D:\Temp\Performancetest>fibs.exe
starting
fib(40)=165580141 took 00:00:03.3648384
fib(40)=165580141 took 00:00:03.4048960

D:\Temp\Performancetest>java convj
starting
convolve=75.88665 took 1.542
convolve=80.14561 took 1.452

D:\Temp\Performancetest>convs.exe
starting
convolve=75,88665 took 00:00:04.5365232
convolve=80,14561 took 00:00:04.5365232
===== Log end =====

It is pretty clear to me which one of these VM:s perform the best...