You don’t seem to have paid attention during OS class. Running at a higher priority does NOT typically increase throughput (frame rate); it only decreases latency in the case where you’re not 100% CPU bound.
You should use a profiler to find out where your program is spending its time, and figure out how to make it not do that. I recommend VTune because it’s so easy to just start up, sample, and then get a view of where the time is spent.
Make sure you profile the release build of your program, not a debug build.
But after the check, what can I do to improve the performace of the program? not by mean of refining the code of OpenGL, but making a better use of the hardware?
if you are unfamiliar with concept of graphics pipeline tuning, I suggest you find some resources on it. Real-time Rendering(2:nd Ed.) by Möller and Haines has a good chapter on it. NVIDIA might also have some papers.
Originally posted by kstsui:
[b]Thx for the advice~
But after the check, what can I do to improve the performace of the program? not by mean of refining the code of OpenGL, but making a better use of the hardware?[/b]
You could start by telling us how you draw your geometry (Immediate Mode, Display Lists, CVA’s, VAR/VAO/VBO?). And the size of your geometry (# of Tris/Verts). Plus what your graphics card is, and if you’ve checked to see if the card is running HW T&L or if it’s fallen back to Software?