FEAR Benchmarking

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

40sTheme

Golden Member
Sep 24, 2006
1,607
0
0
It was an end of review project. We've just finished going over what we learned in regular Probability and Statistics, and this project encompassed all of Statistics. So it was basically just going over experimental design, sampling techniques, etc. However, I told the teacher that I was going to prove it is possible to sample by convenience and get unbiased results. Which I did. :)
 

evolucion8

Platinum Member
Jun 17, 2005
2,867
3
81
I know is too late, I was working from thursday til last night and I didn't have any time to test, here I come.

1024x768
Min 26
AVG 54
Max 143

1280x1024
Min 19
AVG 38
Max 89


These tests were performed twice to make sure the results were final, with 4x FSAA and 16x Aniso with everything on Maxed, except Soft Shadows which looks awful at 1024x768, I could play it higher, but my CRT at 1280x1024 is at 60Hz which is eyestraining, not bad for a almost 3 years videocard based on an almost 5 years architecture, only enhanced. Great testament of the power of the Radeon 9700PRO architecture!!!
 

40sTheme

Golden Member
Sep 24, 2006
1,607
0
0
Yeah, those X800's can still pump it out...
If only they were SM3.0! Nah, they'd run slower then. :)
 

evolucion8

Platinum Member
Jun 17, 2005
2,867
3
81
Originally posted by: 40sTheme
Yeah, those X800's can still pump it out...
If only they were SM3.0! Nah, they'd run slower then. :)

Well, the SM2.0b supported by the X800, is just a SM3.0 without Dynamic Branching. Dynamic Branching allows to execute more than 512 shader instructions on each component, vector, scalar and texture, for a total of 65,536, that differentiate itself from the SM2.0b which only uses static branching allowing only a max of 1,536 shader instruction count. Anyways, no game exceeds the 300 shader count anyways, the performance incurred using dynamic branching or high shader count is just too great for any card right now, that's why DX10 came now, to solve those performance issues when using heavy shader games.

PS: OpenEX HDR is not a SM3.0 feature, and neither the True Displacement Mapping which is a Vertex Shader feature that can be enabled on any ATi DX9 card using the R2VB (Render To Vertex Buffer) command.