Second Pipe and Intersection Geometry
Jeff Brickley (jbrickley++at++lmwsmr.lesc.lockheed.com)
Wed, 04 Jun 1997 09:07:24 -0700
I'll try to rephrase my questions so that they are a little more
specific. I have an idea for using Intersection geometry for finding a
solution to problem normally associated with super-computing. It's rather
an odd use of intersection geometry so I need a little more information
before I risk the time associated with building the program.
I'm using Performer 2.1 and Coryphaeus Software's EasyScene on a 16
processor R10000 Onyx Infinite Reality with 8 RM6s (64MB ea.) on a dual pipe
system. Our main application, which must be running, requires one dedicated
pipe and utilizes 5 forked processes as well as any processes generated by
the Performer mode (APP_CULLDRAW_ISEC) which I believe is 3, leaving me with
8 processors to do some additional work from and the second pipe which is
not used by our primary Real-time display.
So, here are my questions:
1) When exactly does the ISEC process occur and can it's time be
calculated by hand prior to execution? Is there additional
overhead created by the use of intersection geometry other
than the time of the process itself (such as transfer of
information
to the process)?
2) Does the size of the polygons affect calculation of an Intersection
(I have several long skinny triangles)?
3) Can a second Performer program execute assuming that it only
utilizes the second pipe and the remaining processors?
4) If #3 is possible, can I guarantee that the new process will not
interfere in any way with the frame rate speed of the primary
display (the first Performer process)? Does this require a
special setup or will the SGI do this for me?
I can write the new program utilizing EasyScene or as a full Performer only
application, but I need to know these answers to see if my new program is
even feasible.... I am short on time, so I need to get this right the first
time, if I can do it at all, and I cannot risk interfering with our primary
display process....
Thank you for your time.
Jeffry J. Brickley
3D Systems Programmer
Lockheed Martin
White Sands Missile Range, NM
=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
Submissions: info-performer++at++sgi.com
Admin. requests: info-performer-request++at++sgi.com
SGI DevForum 97 info: http://www.sgi.com/Forum97/
This archive was generated by hypermail 2.0b2
on Mon Aug 10 1998 - 17:55:24 PDT