MunkyMark

[Munky]

Well, of course I'll work on it. It really bugs me when my programs dont work like they're supposed to.

But, anyways, I'm starting to think it may have something to do with HT. It's definitely not supposed to take more than a second or two to show the rendering. I'll see if I can come up with some more ideas.

 

[Paratus]

Thanks Munky!

I enjoyed the first Munky Marks - looking forward to this one too!

If I get a chance maybe I'll disable HT and see if that helps.

 

[Sentry2]

Originally posted by: munky
Well, of course I'll work on it. It really bugs me when my programs dont work like they're supposed to.

Maybe you should go work for DICE. They could use someone like you... j/k :beer:

 

[dug777]

I can get it to run perfectly (altho the scores seem very low, and my FBO score is higher than my Sm2 and SM3 scores )...just need task manager to end it...

 

[Munky]

Ok, guys, I've included several test programs with this newest release as of 9/17. They aren't supposed to show anything, just run em to see if you can quit them via the Esc key without crashing. I've managed to run them all successfuly on my anicient P4 1.6 + GF2 MX, so none of your rigs have any excuse for crashing. There are 7 test apps, so let me know which ones can run normally, and which ones crash or refuse to quit.

 

[JRW]

Originally posted by: munky
Ok, guys, I've included several test programs with this newest release as of 9/17. They aren't supposed to show anything, just run em to see if you can quit them via the Esc key without crashing. I've managed to run them all successfuly on my anicient P4 1.6 + GF2 MX, so none of your rigs have any excuse for crashing. There are 7 test apps, so let me know which ones can run normally, and which ones crash or refuse to quit.

None of the test programs crash , im able to exit all of them with the ESC key EXCEPT for test-1.exe , I have to end task on that one.

 

[imported_Truenofan]

e6300 w/ x1900 cf edition, 1gb ram.

sm3: 25
sm2: 15
fbo: 15.2
cpu: 2.59

edit: first time go, had to write down the numbers to post em lol. first time i didnt. now i realize how powerful the x1900 is.

 

[Munky]

Originally posted by: JRW

Originally posted by: munky
Ok, guys, I've included several test programs with this newest release as of 9/17. They aren't supposed to show anything, just run em to see if you can quit them via the Esc key without crashing. I've managed to run them all successfuly on my anicient P4 1.6 + GF2 MX, so none of your rigs have any excuse for crashing. There are 7 test apps, so let me know which ones can run normally, and which ones crash or refuse to quit.

None of the test programs crash , im able to exit all of them with the ESC key EXCEPT for test-1.exe , I have to end task on that one.

Great! Based on those results, I've made some updates to the main application, so run it and see what happens. Also, it now writes a debugging log, so if it crashes, tell me what's in the log.

 

[Munky]

Originally posted by: Truenofan
e6300 w/ x1900 cf edition, 1gb ram.

sm3: 25
sm2: 15
fbo: 15.2
cpu: 2.59

edit: first time go, had to write down the numbers to post em lol. first time i didnt. now i realize how powerful the x1900 is.

Your e6300 isn't overclocked, is it? I get about 2.89 cpu fps on my 2.4ghz Opteron 165. This app doesnt use any SSE/2 optimizations, so it's purely dependent on the FPU performance of the processor. The math loop, though, is written in a way that should allow the cpu to decode and execute several instructions at once.

 

[JRW]

Originally posted by: munky

Originally posted by: JRW

Originally posted by: munky
Ok, guys, I've included several test programs with this newest release as of 9/17. They aren't supposed to show anything, just run em to see if you can quit them via the Esc key without crashing. I've managed to run them all successfuly on my anicient P4 1.6 + GF2 MX, so none of your rigs have any excuse for crashing. There are 7 test apps, so let me know which ones can run normally, and which ones crash or refuse to quit.

None of the test programs crash , im able to exit all of them with the ESC key EXCEPT for test-1.exe , I have to end task on that one.

Great! Based on those results, I've made some updates to the main application, so run it and see what happens. Also, it now writes a debugging log, so if it crashes, tell me what's in the log.

Now when I hit ESC the program crashes (first time thats happened) heres what the log shows:

--- Mandelbrot Shader Info Log ---
Renderer: GeForce 7800 GTX/PCI/SSE2/3DNOW!

Extensions: GL_ARB_color_buffer_float GL_ARB_depth_texture GL_ARB_draw_buffers GL_ARB_fragment_program GL_ARB_fragment_program_shadow GL_ARB_fragment_shader GL_ARB_half_float_pixel GL_ARB_imaging GL_ARB_multisample GL_ARB_multitexture GL_ARB_occlusion_query GL_ARB_pixel_buffer_object GL_ARB_point_parameters GL_ARB_point_sprite GL_ARB_shadow GL_ARB_shader_objects GL_ARB_shading_language_100 GL_ARB_texture_border_clamp GL_ARB_texture_compression GL_ARB_texture_cube_map GL_ARB_texture_env_add GL_ARB_texture_env_combine GL_ARB_texture_env_dot3 GL_ARB_texture_float GL_ARB_texture_mirrored_repeat GL_ARB_texture_non_power_of_two GL_ARB_texture_rectangle GL_ARB_transpose_matrix GL_ARB_vertex_buffer_object GL_ARB_vertex_program GL_ARB_vertex_shader GL_ARB_window_pos GL_ATI_draw_buffers GL_ATI_texture_float GL_ATI_texture_mirror_once GL_S3_s3tc GL_EXT_texture_env_add GL_EXT_abgr GL_EXT_bgra GL_EXT_blend_color GL_EXT_blend_equation_separate GL_EXT_blend_func_separate GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_compiled_vertex_array GL_EXT_Cg_shader GL_EXT_depth_bounds_test GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_framebuffer_object GL_EXT_multi_draw_arrays GL_EXT_packed_depth_stencil GL_EXT_packed_pixels GL_EXT_pixel_buffer_object GL_EXT_point_parameters GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_separate_specular_color GL_EXT_shadow_funcs GL_EXT_stencil_two_side GL_EXT_stencil_wrap GL_EXT_texture3D GL_EXT_texture_compression_s3tc GL_EXT_texture_cube_map GL_EXT_texture_edge_clamp GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_EXT_texture_filter_anisotropic GL_EXT_texture_lod GL_EXT_texture_lod_bias GL_EXT_texture_mirror_clamp GL_EXT_texture_object GL_EXT_texture_sRGB GL_EXT_timer_query GL_EXT_vertex_array GL_HP_occlusion_test GL_IBM_rasterpos_clip GL_IBM_texture_mirrored_repeat GL_KTX_buffer_region GL_NV_blend_square GL_NV_copy_depth_to_color GL_NV_depth_clamp GL_NV_fence GL_NV_float_buffer GL_NV_fog_distance GL_NV_fragment_program GL_NV_fragment_program_option GL_NV_fragment_program2 GL_NV_half_float GL_NV_light_max_exponent GL_NV_multisample_filter_hint GL_NV_occlusion_query GL_NV_packed_depth_stencil GL_NV_pixel_data_range GL_NV_point_sprite GL_NV_primitive_restart GL_NV_register_combiners GL_NV_register_combiners2 GL_NV_texgen_reflection GL_NV_texture_compression_vtc GL_NV_texture_env_combine4 GL_NV_texture_expand_normal GL_NV_texture_rectangle GL_NV_texture_shader GL_NV_texture_shader2 GL_NV_texture_shader3 GL_NV_vertex_array_range GL_NV_vertex_array_range2 GL_NV_vertex_program GL_NV_vertex_program1_1 GL_NV_vertex_program2 GL_NV_vertex_program2_option GL_NV_vertex_program3 GL_NVX_conditional_render GL_SGIS_generate_mipmap GL_SGIS_texture_lod GL_SGIX_depth_texture GL_SGIX_shadow GL_SUN_slice_accum GL_WIN_swap_hint WGL_EXT_swap_control

number of threads: 2
shader version: 3
support multitexturing: 1
support FBO: 1
fontlist: 1
quadlist: 97
fliplist: 98
palette texture id: 1
FBO id: 1
pingpong Tex IDs: [2, 3]
FBO color attachments: [8CE0, 8CE1]
vertex program: 1
fragment programs: [2, 3, 4, 5, 6]
program objects: [7, 8, 9, A, B]

 

[imported_Truenofan]

nah, this board doesnt oc well. so i just left it stock speeds. i plan on getting a asus p5w dh deluxe and overclocking.

edit: this is my first intel build, and im actualy loving it. cant wait to extract the most out of this thing, judging by how well it oc's on air, i plan on going to water, and possibly hitting closer to 4ghz.

 

[Paratus]

All the tests worked - the ESC key would let me out - however the main app still hangs.

Here's the debug log:

--- Mandelbrot Shader Info Log ---
Renderer: RADEON 9600 XT x86/SSE2

Extensions: GL_ARB_multitexture GL_EXT_texture_env_add GL_EXT_compiled_vertex_array GL_S3_s3tc GL_ARB_depth_texture GL_ARB_fragment_program GL_ARB_fragment_program_shadow GL_ARB_fragment_shader GL_ARB_multisample GL_ARB_occlusion_query GL_ARB_point_parameters GL_ARB_point_sprite GL_ARB_shader_objects GL_ARB_shading_language_100 GL_ARB_shadow GL_ARB_shadow_ambient GL_ARB_texture_border_clamp GL_ARB_texture_compression GL_ARB_texture_cube_map GL_ARB_texture_env_add GL_ARB_texture_env_combine GL_ARB_texture_env_crossbar GL_ARB_texture_env_dot3 GL_ARB_texture_mirrored_repeat GL_ARB_transpose_matrix GL_ARB_vertex_blend GL_ARB_vertex_buffer_object GL_ARB_vertex_program GL_ARB_vertex_shader GL_ARB_window_pos GL_ARB_draw_buffers GL_ATI_draw_buffers GL_ATI_element_array GL_ATI_envmap_bumpmap GL_ATI_fragment_shader GL_ATI_map_object_buffer GL_ATI_separate_stencil GL_ATI_texture_env_combine3 GL_ATI_texture_float GL_ATI_texture_mirror_once GL_ATI_vertex_array_object GL_ATI_vertex_attrib_array_object GL_ATI_vertex_streams GL_ATIX_texture_env_combine3 GL_ATIX_texture_env_route GL_ATIX_vertex_shader_output_point_size GL_EXT_abgr GL_EXT_bgra GL_EXT_blend_color GL_EXT_blend_func_separate GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_clip_volume_hint GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_framebuffer_object GL_EXT_multi_draw_arrays GL_EXT_packed_pixels GL_EXT_point_parameters GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_separate_specular_color GL_EXT_shadow_funcs GL_EXT_stencil_wrap GL_EXT_texgen_reflection GL_EXT_texture3D GL_EXT_texture_compression_s3tc GL_EXT_texture_cube_map GL_EXT_texture_edge_clamp GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_EXT_texture_filter_anisotropic GL_EXT_texture_lod_bias GL_EXT_texture_mirror_clamp GL_EXT_texture_object GL_EXT_texture_rectangle GL_EXT_vertex_array GL_EXT_vertex_shader GL_HP_occlusion_test GL_NV_blend_square GL_NV_occlusion_query GL_NV_texgen_reflection GL_SGI_color_matrix GL_SGIS_generate_mipmap GL_SGIS_multitexture GL_SGIS_texture_border_clamp GL_SGIS_texture_edge_clamp GL_SGIS_texture_lod GL_SUN_multi_draw_arrays GL_WIN_swap_hint WGL_EXT_extensions_string WGL_EXT_swap_control

number of threads: 2
shader version: 2
support multitexturing: 1
support FBO: 1
fontlist: 1
quadlist: 97
fliplist: 98
palette texture id: 1
FBO id: 1
pingpong Tex IDs: [2, 3]
FBO color attachments: [8CE0, 8CE1]
vertex program: 40000001
fragment programs: [20000001, 20000002, 20000003, 20000004, 20000005]
program objects: [80000001, 80000002, 80000003, 80000004, 80000005]

 

[nullpointerus]

Originally posted by: dug777
Bets i get the lowest scores

2.1Ghz tbred-b, 6600GT@585/1145, 1GB RAM

SM3 = 2.625
SM2 = 2
FBO = 3
CPU = 1.166667

Heh, almost. My 7900GT got shipped back under RMA, so I'm using an old 6200TC.

1.8GHz Athlon 64 (939), 6200TC@550/700, 2GB RAM

SM3 = 1.13
SM2 = 1.14
FBO = 1.00
CPU = 1.14

 

[Munky]

Hmmm.... well, I'll have to look into it and see what else I can change to make it work. Incidentally, the main app also crashed on my gf2 mx, so I got more work to do.

 

[dug777]

Originally posted by: nullpointerus

Originally posted by: dug777
Bets i get the lowest scores

2.1Ghz tbred-b, 6600GT@585/1145, 1GB RAM

SM3 = 2.625
SM2 = 2
FBO = 3
CPU = 1.166667

Heh, almost. My 7900GT got shipped back under RMA, so I'm using an old 6200TC.

1.8GHz Athlon 64 (939), 6200TC@550/700, 2GB RAM

SM3 = 1.13
SM2 = 1.14
FBO = 1.00
CPU = 1.14

i wonder why my FBO score is higher than my hardware scores?

 

[josh6079]

Sticky? This would be nice to serve as a forum wide benchmark for video cards. It has my vote.

 

[Munky]

Originally posted by: dug777

Originally posted by: nullpointerus

Originally posted by: dug777
Bets i get the lowest scores

2.1Ghz tbred-b, 6600GT@585/1145, 1GB RAM

SM3 = 2.625
SM2 = 2
FBO = 3
CPU = 1.166667

Heh, almost. My 7900GT got shipped back under RMA, so I'm using an old 6200TC.

1.8GHz Athlon 64 (939), 6200TC@550/700, 2GB RAM

SM3 = 1.13
SM2 = 1.14
FBO = 1.00
CPU = 1.14

i wonder why my FBO score is higher than my hardware scores?

My friend's 7800gt also got a higher FBO score. The FBO method also used PS 2.0, except it does not run all 255 iterations in one loop. Every 5 iterations it stores the result in a texture, and it reads the result back at the beginning of the next 5 iterations. On a card like the x1900xt it will arguably be limited by texturing performance since it has about 3x as much shaders as texture units. My older version had a texture read/write every single iteration, and it ran at like 6 fps on a x1900xt. On a 6600gt, you have as many shaders as texture units, and the driver probably can compile the shader into shorter code, so it may be running more efficiently on a 6600gt than the single loop shader.

 

[dug777]

Originally posted by: munky

Originally posted by: dug777

Originally posted by: nullpointerus

Originally posted by: dug777
Bets i get the lowest scores

2.1Ghz tbred-b, 6600GT@585/1145, 1GB RAM

SM3 = 2.625
SM2 = 2
FBO = 3
CPU = 1.166667

Heh, almost. My 7900GT got shipped back under RMA, so I'm using an old 6200TC.

1.8GHz Athlon 64 (939), 6200TC@550/700, 2GB RAM

SM3 = 1.13
SM2 = 1.14
FBO = 1.00
CPU = 1.14

i wonder why my FBO score is higher than my hardware scores?

My friend's 7800gt also got a higher FBO score. The FBO method also used PS 2.0, except it does not run all 255 iterations in one loop. Every 5 iterations it stores the result in a texture, and it reads the result back at the beginning of the next 5 iterations. On a card like the x1900xt it will arguably be limited by texturing performance since it has about 3x as much shaders as texture units. My older version had a texture read/write every single iteration, and it ran at like 6 fps on a x1900xt. On a 6600gt, you have as many shaders as texture units, and the driver probably can compile the shader into shorter code, so it may be running more efficiently on a 6600gt than the single loop shader.

Interesting stuff

 

[nullpointerus]

Originally posted by: nullpointerus

Originally posted by: dug777
Bets i get the lowest scores

2.1Ghz tbred-b, 6600GT@585/1145, 1GB RAM

SM3 = 2.625
SM2 = 2
FBO = 3
CPU = 1.166667

Heh, almost. My 7900GT got shipped back under RMA, so I'm using an old 6200TC.

1.8GHz Athlon 64 (939), 6200TC@550/700, 2GB RAM

SM3 = 1.13
SM2 = 1.14
FBO = 1.00
CPU = 1.14

LOL, misread the clock speeds on my card. They should be 350/550.

 

[Munky]

All the tests worked - the ESC key would let me out - however the main app still hangs.

Ok, beta 3 is available for downloading. I added a second version of the main app that will initialize everything as normal, but will not actually render anything. That way, if it crashes I'll know if it hangs while loading or while rendering. Also, the debug log now writes what version of OpenGL your video drivers support. Let me know if it's lower than 2.0

 

[JRW]

Both of the latest .exe's exit fine for me now without crashing :thumbsup:
OpenGL version: 2.0.3

 

[xtknight]

(Bolded portion may have something to do with it.) Instead of ending gracefully (on startup with limited OpenGL setup) it ends with this (this is a window on top of a black viewport).

Also, with software rendering, the ESC key should probably be run in a separate thread (it takes about 5 seconds for the ESC to register in software mode). Preferably instead of exiting it should permit only software mode. Other than that, looking great and I look forward to see what becomes of this. It's still working great in wine/Linux so major props for that. Finally a more or less cross-platform benchmark.

--- Mandelbrot Shader Info Log ---
MakeGLWindow: registered class
MakeGLWindow: fullscreen mode successful
SizeGLScreen: set projection
SizeGLScreen: set viewport
MakeGLWindow: created window
InitGL: acquired device context
InitGL: setup pixel format
InitGL: created rendering context
InitGL: bound rendering context
InitGL: set ogl parmeters
InitGL: checked extensions
SizeGLScreen: set projection
SizeGLScreen: set viewport
CreateBitmapFont: font created
Init: detected 1 processors
LoadBMP: load completed
Init: allocated imgBuffer
Init: allocated thread handles
Init: created mutex handle
Init: generated quad list
Init: generated flip list
Init: compiled quadlist
Init: compiled fliplist
LoadBMP: load completed
CreateTexture: creation complete
Init: shader version not > 1
Renderer: GDI Generic
OpenGL version: 1.1.0

Extensions: GL_WIN_swap_hint GL_EXT_bgra GL_EXT_paletted_texture

number of threads: 1
shader version: 0
support multitexturing: 0
support FBO: 0
fontlist: 1
quadlist: 97
fliplist: 98
palette texture id: 1
FBO id: 0
pingpong Tex IDs: [0, 0]
FBO color attachments: [8CE0, 8CE1]
vertex program: 0
fragment programs: [0, 0, 0, 0, 0]
program objects: [0, 0, 0, 0, 0]
Init: reset view
Cleanup: closed mutex handle
Cleanup: released thread handles
Cleanup: freed palette data
Cleanup: freed imgBuffer data
Cleanup: deleted palette texture
Cleanup: deleted quadlist
Cleanup: deleted fliplist
Cleanup: deleted fontlist
KillGLWindow: unbind and delete rendering context
KillGLWindow: release device context
SizeGLScreen: set projection
SizeGLScreen: set viewport
KillGLWindow: exit fullscreen mode
KillGLWindow: destroyed window
KillGLWindow: unregistered class

 

[Bill Kunert]

A64 3500+ @ 2.4GHz
1.5 Gig PC3200
7300GT 128Meg DDR3

SM3=3.44
SM2=3.10
FBO=3.88
CPU=1.63

Bill

 

[Munky]

Originally posted by: xtknight
(Bolded portion may have something to do with it.) Instead of ending gracefully (on startup with limited OpenGL setup) it ends with this (this is a window on top of a black viewport).

Also, with software rendering, the ESC key should probably be run in a separate thread (it takes about 5 seconds for the ESC to register in software mode). Preferably instead of exiting it should permit only software mode. Other than that, looking great and I look forward to see what becomes of this. It's still working great in wine/Linux so major props for that. Finally a more or less cross-platform benchmark.

I got the same results when I had no graphics driver installed on the gf2 mx. When it says "GDI Generic" for renderer, it means that OpenGL is emulated in software on the system. What I think is causing the crash is that even though there's a cpu math mode, the app still uses OpenGL to draw the result buffer, and with limited/incomplete OGL support it causes an error. When it says "shader version not > 1", it means SM2 or higher shaders are not supported in HW, and the app will only run using the cpu mode. I may just have to code a pure software rendering mode that uses no OpenGL at all.

*edit: BTW, does the idle version of the app crash also?

 

[Munky]

And in other news, I just ran this on a rig with a e6600 at stock 2.4ghz, and it is getting the same cpu score as my 2.4ghz Opteron 165. Either Intel did not improve the FPU units much, or I gotta rewrite the cpu math loop to allow for more instruction per clock.