| >Burton Alexandre wrote:
>>
>> anyone interested in a "serious" csound benchmarking effort?
Sure... I was curious what sort of difference I'd get by comparing dos
csound/win95 csound/linux csound csound/linux & X csound... I tried
running the currently available benchmarks and everything went so fast I
wasn't sure what, if anything, to conclude.
Another suggestion: I wouldn't mind some benchmarks for testing realtime
capacity... RTCmix (cmix hacked to give realtime output) has a little
benchmark (? more appropriate term for realtime output??) test that just
adds oscillators one at a time until your output gets bad...
so your benchmark score is the number of successful simultaneous
oscillators you got.
See http://www.panix.com/~topper/Cmix/benchmarks.html
They also have links to an experimental compiler that allows pentium
optimization-- this might be useful to folks trying to squeeze more
performance out of linux csound on pentium boxes. For RTCmix the
reported realtime performance increases range from about 9% to about 20
% ... the question, of course, is whether you can get it work without
intolerable amounts of instability... I haven't tried it (yet).
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com |