On Fri, Jan 27, 2012 at 9:01 PM, Robert Osfield <robert.osfi...@gmail.com> wrote: ... > Curiously running separate contexts via separate applications has in > the past shown good scalability so I suspect the driver contention is > something per application specific. > > Along these lines I do wonder if running separate processes for each > context rather than separate threads might help avoid the contention. > Unfortunately the OSG's osgViewer design is based around the > lightweight data sharing that threading provides over use of separate > processes, and OpenThreads is geared up for creation and management of > threads rather than processes. >
I looked into this a bit and couldn't conclude that running in separate processes makes much of a difference. The best I could say is that the applications would sometimes show the expected (i.e., single screen) frame rate, but would then fall into periods of the lower frame rate seen in the multi-threaded case. I don't have anything genlocked, so the programs weren't synchronized. I believe that the high rates were achieved when the programs happened to be making their OpenGL calls at different times, but then due to a scheduling artifact they started overlapping their graphics calls. Tim _______________________________________________ osg-users mailing list osg-users@lists.openscenegraph.org http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
