Csound Csound-dev Csound-tekno Search About

PPC Benchmarks

Date1997-04-23 17:55
FromGrant Chu Covell
SubjectPPC Benchmarks
On a PowerComputing PowerTower 180e (MacOs 7.5.5, Mills Csound PPC
0.2.0b):

bach -d	bach -m	riss -g	riss -m	guit -d	guit -m	jame -g pvanal	lpanal
17.70	7.18	8.99	3.65	16.15	5.45	0.68	9.16	2.02

I always knew messages off was faster, just never thought it could be
this much faster: Without any visual indication I always thought the
process had died. There are faster PPCs out there now; this 180e (the e
is for 604e) replaced a IIsi, and I can imagine that this 180e will soon
be the IIsi of the future!!

Grant
gcc@kodak.com

Date1997-04-23 19:23
FromDavid Madole
SubjectRe: PPC Benchmarks
> 
> On a PowerComputing PowerTower 180e (MacOs 7.5.5, Mills Csound PPC
[...]
> I always knew messages off was faster, just never thought it could be
> this much faster: Without any visual indication I always thought the
> process had died. There are faster PPCs out there now; this 180e (the e

A good compromise is to set the message buffer to a large value.  This
allows you to still see the messages without the program doing a lot
of I/O.  The difference suprised us too.

You should ALWAYS set messages off for real-time MIDI, though.

We have put some strategically placed buffer flushes in the upcoming
(really, real soon now) release that should fix the "thought the
process had died" problem.

Dave

Dave Madole
Technical Director, Center for Contemporary Music
Listserv Administrator

Mills College
Oakland, CA 94613
510-430-2336

madole@mills.edu