Diamond FireGL2

r2k

Distinguished
Jan 27, 2001
414
0
18,780
It'd be very kind of a Tom's Hardware Team Member to answer my questions about their review of Diamond FireGL2 video card and their tests on 3D Studio MAX R3:
I have a Celeron 466 o/c to 546 MHz with a very old video card(which I'm going to replace as soon as I earn something!) Trident 3Dimage975.My friend has a P-III 550 o/c to 605 MHz and a GeForce2-MX.His system is much much faster than mine in the Modelling and shading process(when we both build the same models,shade them,move them,texture them...) but the difference in the actual rendering time is NEGLIGIBLE meaning 3D Studio is not benefiting from his better video card.I'm stuck with the Software Z-Buffer setting in the display driver setting and he's used both OpenGL and DirectX with comparable results.
So I wanted to know:
1.I suppose you used the OpenGL driver setting(in 3DStudio;I mean the default OpenGL setting) or was it any custom driver especially optimized?
2.Were your tests RENDERING of the animations or simply doing things in the main viewports,shading,etc.?
3.What was the resolution you rendered at?(I mean the 3D Studio settings like 640*480,320*200,....)
4.How did you get the results?(for example 5.02 frames/s)Did you use a stopwatch(?!!) or some other internal numbers shown to you by MAX?

Thank you very much for your great articles!
 
G

Guest

Guest
i don't believe that the graphics card has any effect on rendering time... that is solely determined by the cpu(s). the graphics card is only used for modeling.
 

r2k

Distinguished
Jan 27, 2001
414
0
18,780
I thought so;But then tell me what are the numbers produced in the R3 benchmark tests?How did they come up with them?
I think this thing deserves an answer from Tom's Hardware team to clear up,don't you think so?
 
G

Guest

Guest
i think it briefly mentions that the benchmark program has camera move around the scene and depending on the speed of the card the camera will be able to give more or less frames per second (in the viewport). so it's a messure of how smoothly you can navigate your scene depending on complexity and which features are enabled.
 

r2k

Distinguished
Jan 27, 2001
414
0
18,780
Thanks for the good info!But how do you EXACTLY measure the Frames/Second?I haven't seen a 3DStudio MAX R3.1 feature about this...Are they using the default animation of the model?If that's so,maybe they're dividing the number of animation frames in each benchmark file to the time 3D Studio finishes the whole animation. They probably animate the scene many times to reach the overall FPS...
I think this is getting clearer though,again thanks DeSilentio!