Simon Dimaio (simond++at++ee.ubc.ca)
Thu, 29 May 1997 12:24:17 -0700 (PDT)
This is a question regarding graphics performance. I have written a
relatively small visual simulation on an old PowerSeries system with 2
processors (R2000A/R3000) and VGX graphics, which is to be ported to an O2
with an R5000.
Channel statistics (Powerseries) show the following process times:
Application = 4.6ms
Cull = 4ms
Draw = 40ms
Yet the total frame time is reported as 75ms.
After the addition of some "on the fly" geometry (Geode), the process
times become.
Application = 4.6ms
Cull = 11ms
Draw = 60ms
Total frame time = 120ms!
On the Powerseries, locked at 60Hz, the graphics looks reasonable. The O2
is not capable of locking the frame rate and the simulation becomes
extremely sluggish.
Why is the frame time so much longer than the sum of APP+CULL+DRAW?
Surely, in APPCULL_DRAW multiprocess mode, my frame time should be more
like 70ms?
I know that this is a tricky question to answer without seeing the code,
but how can I optimise my simulation? I have used backface culling
wherever possible, have flattened the scene tree and have put all the
non-latency critical simulation updates after pfFrame().
Thanks for your time.
Regards
Simon Dimaio
=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
Submissions: info-performer++at++sgi.com
Admin. requests: info-performer-request++at++sgi.com
This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:55:19 PDT