| Title: | open3d |
| Notice: | Kits on notes 3 and 4; Documents note 223 |
| Moderator: | WRKSYS::COULTER |
| Created: | Wed Dec 09 1992 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 1306 |
| Total number of notes: | 5260 |
Hi all, found on the www a version of viewperf which runs on Digital UNIX: avalon.eng.pko.dec.com:80/open3d/docs/opengl/performance/opc/viewperf.5.0.tar (no source, executable only) Can we give this version to a customer, who wants to verify the OpenGL performance numbers of his PowerStorm 4D20 ? This customer claims the OpenGL performance of his 4D20 is 10-20 times lower than the specs say. Either his own OpenGL programs are programmed bad or the card is slow. /Reinhard
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 1238.1 | WRKSYS::CHALTAS | Never trust a talking mime | Wed Feb 05 1997 10:55 | 18 | |
Real-world programs rarely match the performance level of industry
standard benchmarks, since they have the annoying habit of trying to
solve real-world problems at the same time.
The customer should use the viewperf numbers as a general guide to the
performance potential of the graphics card, not a target which his
application (unless it is *also* viewperf) should be excected to
acheive.
There are any number of reasons why an application may not drive a
graphics card to its fullest potential -- far too many to list, but in
general performance is affected by the type and complexity of the
primitives being rendered (for instance, long triangle strips are
faster than discrete triangles), the lighting algorithms (if any -- no
lights at all is usually fastest) being used, and even the size of the
window being displayed.
George
| |||||