Hello,
I'm planning to update the benchmarks on the wiki with version 9.20 when it's released and thought I would ask in advance if anyone here has any ideas about the benchmarks. The three new benchmarks that I plan to add are 3DMark03, 3DMark05?? and 3DMark06
Anyone here interested in ShaderMark or AquaMark? or any other bench for that matter.
Also Ive made a small update to : http://wiki.winehq.org/BenchMark-0%2e9%2e6 The test were not run under optimal conditions, other apps open, not from a clean boot, run in a window, and with debugging on... So the scores should improve :)
Cheers,
Tom
On 8/1/06, Tom Wickline twickline@gmail.com wrote:
Hello,
I'm planning to update the benchmarks on the wiki with version 9.20 when it's released and thought I would ask in advance if anyone here has any ideas about the benchmarks. The three new benchmarks that I plan to add are 3DMark03, 3DMark05?? and 3DMark06
Anyone here interested in ShaderMark or AquaMark? or any other bench for that matter.
It would be gr8 to evaluate wine's shader capabiliies with windows.
Also Ive made a small update to : http://wiki.winehq.org/BenchMark-0%2e9%2e6 The test were not run under optimal conditions, other apps open, not from a clean boot, run in a window, and with debugging on... So the scores should improve :)
Cheers,
Tom
It might be a good idea to fix that 3DMark03 performance regression caused by VBOs :-)
On Tuesday, August 01, 2006 10:39, H. Verbeet wrote:
It might be a good idea to fix that 3DMark03 performance regression caused by VBOs :-)
Yes, http://bugs.winehq.org/show_bug.cgi?id=5546 for reference.
As for tests #3 and #4 in 3DMark 03, they both run, although 3DMark 03 frequently crashes while loading test #3. Test #4 requires using GLSL instead of ARB shaders though.
As for 3DMark 05 and 06, they don't seem to think we support Pixel Shader 2.0, even when using GLSL, so I'm not sure how valuable those would be to benchmark in that state.
- Neil
On 8/1/06, Neil Skrypuch ns03ja@brocku.ca wrote:
As for tests #3 and #4 in 3DMark 03, they both run, although 3DMark 03 frequently crashes while loading test #3. Test #4 requires using GLSL instead of ARB shaders though.
Yea, Ive found this to be the case.. I have however got test #3 to run without GLSL the only problem is its rather crash prone as you have pointed out.
As for 3DMark 05 and 06, they don't seem to think we support Pixel Shader 2.0, even when using GLSL, so I'm not sure how valuable those would be to benchmark in that state.
That's why I put the ?? after 05... As for 06 ive got it running, well if that's what you want to call it.. all the supported test will run at 800x600 accept for #2 (Firefly) it will run at 512x384 and 3dmark06 gives a score. :-)
I uploaded a shot: http://wiki.winehq.org/BenchMark-0%2e9%2e6
Keep in mind all but one of the test are rendered incorrectly at the moment..
Tom
- Neil
On 02/08/06, Tom Wickline twickline@gmail.com wrote:
As for 3DMark 05 and 06, they don't seem to think we support Pixel Shader 2.0, even when using GLSL, so I'm not sure how valuable those would be to benchmark in that state.
I've come accross some sample programs that seem to use a reference device to check what shader versions we support. I don't know if 3DMark05/06 do that as well, but it might be something to look for. +d3d_caps should be enough to tell.
On 8/2/06, H. Verbeet hverbeet@gmail.com wrote:
On 02/08/06, Tom Wickline twickline@gmail.com wrote:
As for 3DMark 05 and 06, they don't seem to think we support Pixel Shader 2.0, even when using GLSL, so I'm not sure how valuable those would be to benchmark in that state.
I've come accross some sample programs that seem to use a reference device to check what shader versions we support. I don't know if 3DMark05/06 do that as well, but it might be something to look for. +d3d_caps should be enough to tell.
There's also a chance that some programs check the VS20CAPS or PS20CAPS structures which we currently don't fill at all, and report 0's. These should be fairly easy to fill as they don't look that complicated from MSDN.
On 8/1/06, H. Verbeet hverbeet@gmail.com wrote:
It might be a good idea to fix that 3DMark03 performance regression caused by VBOs :-)
:-)
On 8/1/06, H. Verbeet hverbeet@gmail.com wrote:
It might be a good idea to fix that 3DMark03 performance regression caused by VBOs :-)
I've done some tweaking to my box and the scores in 3dmark2000, 2001se, 2003 have improved considerably to say the least :-D I uploaded two shots, 2001se and 2003 against 9.19 a couple minutes ago if anyone is interested in the result change from 9.18 to 9.19 just scroll down the page to see 9.18 results.
http://wiki.winehq.org/BenchMark-0%2e9%2e6
I still need to do some more tweaking tho, and as before these test are not run under optimal conditions. 3dmark2001se should be over 10,000 when everything is ready to go.
Tom
On 16/08/06, Tom Wickline twickline@gmail.com wrote:
I still need to do some more tweaking tho, and as before these test are not run under optimal conditions. 3dmark2001se should be over 10,000 when everything is ready to go.
You should be able to get the Mother Nature test to run by enabling GLSL. (HKCU/Software/Wine/Direct3D/UseGLSL = enabled)
On 8/16/06, H. Verbeet hverbeet@gmail.com wrote:
On 16/08/06, Tom Wickline twickline@gmail.com wrote:
I still need to do some more tweaking tho, and as before these test are not run under optimal conditions. 3dmark2001se should be over 10,000 when everything is ready to go.
You should be able to get the Mother Nature test to run by enabling GLSL. (HKCU/Software/Wine/Direct3D/UseGLSL = enabled)
:-)
updated with GLSL enabled, left the one with it disabled as well.. Ive got all the test in 2000, 2001se, 2003 running at 1024x768x24
http://wiki.winehq.org/BenchMark-0%2e9%2e6
Almost time to put the screws to it..
On 8/16/06, Tom Wickline twickline@gmail.com wrote:
http://wiki.winehq.org/BenchMark-0%2e9%2e6
Almost time to put the screws to it..
511% increase in 3dmark06 over 0.9.18 What is most interesting is the Perlin Noise [SM 3.0] test shows as being supported and runs, well it tries to run.. :D
-
H. Verbeet -
Jason Green -
Neil Skrypuch -
Tom Wickline -
Vijay Kiran Kamuju