Scott Angster (sangster++at++cme.nist.gov)
Mon, 3 Feb 97 09:03:06 EST
We are working on an immersive virtual assembly tool using Performer 2.0 and Irix 5.3 (haven't upgraded yet), on both an older Crimson RE and an Onyx RE2. In order to keep down the lag, we have put the calls to our tracking devices (Flock of Birds) in separate code that just writes to a shared memory location. The Performer code then just reads out of the shared memory location to get the latest data. This cuts out about 100ms of app time doing it this way. The code that has been developed runs fine on the single processor Crimson but has problems on the Onyx which as two processors. If we force the Onyx onto one processor using pfMultiProc(PFMP_APPCULLDRAW) and pfRunProcOn(0), we get better results but still not the same as a single processor machine. I assume that this is probably due to Performer's handling of its own shared memory arena and that the other shared memory for use with the birds is outside the arena. Can anyone suggest the best solution for this? Some were made in using Sproc but hav
en't had much luck there yet.
Thanks.
Scott
sangster++at++cme.nist.gov
=======================================================================
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 - 17:54:35 PDT