Own an ATI X1, HD or NV 7/8 series GPU and have 5-10 minutes? Please run this.

Discussion in 'GPGPU Technology & Programming' started by PeterT, May 29, 2007.

  1. PeterT

    Regular

    Joined:
    May 14, 2002
    Messages:
    702
    Likes Received:
    14
    Location:
    Austria
    Interesting. The 2 Radeon results from XP seem to confirm my suspicion that the issue only appears on Vista. Now it would be nice to have someone with HD 2900XT and XP run the app. But probably everyone with such cutting-edge HW also runs the bleeding-edge OS. (Oh I'm killing myself)

    It's completely GPU limited - or at least the parallelizable part of it is - so that wouldn't help much. I have a very experimental GPU/CPU combined solver, but that also only runs the very coarse grid levels on the CPU. All the parts that are easily parallelized are faster on GPUs anyway.
     
  2. Geeforcer

    Geeforcer Harmlessly Evil
    Veteran

    Joined:
    Feb 6, 2002
    Messages:
    2,297
    Likes Received:
    464
    LOL, some of us are still pretty happy in the XP land.

    I see... I was just asking because it drove CPU utilization on core 0 to 100%.
     
  3. banksie

    Newcomer

    Joined:
    Jun 9, 2003
    Messages:
    213
    Likes Received:
    4
    Location:
    Wellington, New Zealand
    I'll try running it on my XP machine with a 2900 tonight when I get home from work. Going to be interesting as I have been having odd stability issues and I am wondering if my PSU isn't holding up....

    This should test it nicely.
     
  4. Anarchist4000

    Veteran Regular

    Joined:
    May 8, 2004
    Messages:
    1,439
    Likes Received:
    359
    Not sure how good these results are. Might have interfered with the benchmarking a bit.

    1600XT
    WindowsXP

    Code:
    ------------------ 
     -- size: 3 
      
     BENCHMARK 
     Jacobi iteration: 4.6 micros 
     Residual calculation: 6.3 micros 
     Restriction: fw: 4.7 micros 
     Interpolation + add: 4.7 micros 
     VCycle: 562 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 4.6 micros 
     Residual calculation: 4.7 micros 
     Restriction: fw: 6.3 micros 
     Interpolation + add: 4.7 micros 
     VCycle: 547 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 6.2 micros 
     Residual calculation: 4.7 micros 
     Restriction: fw: 4.7 micros 
     Interpolation + add: 4.7 micros 
     VCycle: 563 micros 
      
      
      
     ------------------ 
     -- size: 7 
      
     BENCHMARK 
     Jacobi iteration: 5.11111 micros 
     Residual calculation: 5.22222 micros 
     Restriction: fw: 5.22222 micros 
     Interpolation + add: 5.22222 micros 
     VCycle: 937.778 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 3.44444 micros 
     Residual calculation: 5.22222 micros 
     Restriction: fw: 5.22222 micros 
     Interpolation + add: 5.22222 micros 
     VCycle: 972.222 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 5.22222 micros 
     Residual calculation: 5.22222 micros 
     Restriction: fw: 5.22222 micros 
     Interpolation + add: 5.22222 micros 
     VCycle: 920 micros 
      
      
      
     ------------------ 
     -- size: 15 
      
     BENCHMARK 
     Jacobi iteration: 5.875 micros 
     Residual calculation: 5.875 micros 
     Restriction: fw: 3.875 micros 
     Interpolation + add: 5.875 micros 
     VCycle: 1406.25 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 5.875 micros 
     Residual calculation: 7.75 micros 
     Restriction: fw: 5.875 micros 
     Interpolation + add: 3.875 micros 
     VCycle: 1308.75 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 5.875 micros 
     Residual calculation: 5.875 micros 
     Restriction: fw: 3.875 micros 
     Interpolation + add: 5.875 micros 
     VCycle: 1347.5 micros 
      
      
      
     ------------------ 
     -- size: 31 
      
     BENCHMARK 
     Jacobi iteration: 11.1429 micros 
     Residual calculation: 11.1429 micros 
     Restriction: fw: 13.4286 micros 
     Interpolation + add: 6.71429 micros 
     VCycle: 1718.57 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 13.2857 micros 
     Residual calculation: 11.1429 micros 
     Restriction: fw: 11.2857 micros 
     Interpolation + add: 11.1429 micros 
     VCycle: 1762.86 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 11.1429 micros 
     Residual calculation: 11.1429 micros 
     Restriction: fw: 13.4286 micros 
     Interpolation + add: 6.71429 micros 
     VCycle: 1718.57 micros 
      
      
      
     ------------------ 
     -- size: 63 
      
     BENCHMARK 
     Jacobi iteration: 44.3333 micros 
     Residual calculation: 44.1667 micros 
     Restriction: fw: 47 micros 
     Interpolation + add: 41.6667 micros 
     VCycle: 2213.33 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 44.1667 micros 
     Residual calculation: 44.3333 micros 
     Restriction: fw: 46.8333 micros 
     Interpolation + add: 28.6667 micros 
     VCycle: 2318.33 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 44.3333 micros 
     Residual calculation: 41.6667 micros 
     Restriction: fw: 49.5 micros 
     Interpolation + add: 28.6667 micros 
     VCycle: 2370 micros 
      
      
      
     ------------------ 
     -- size: 127 
      
     BENCHMARK 
     Jacobi iteration: 334.2 micros 
     Residual calculation: 3.2 micros 
     Restriction: fw: 190.6 micros 
     Interpolation + add: 112.6 micros 
     VCycle: 3874 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 162.6 micros 
     Residual calculation: 171.8 micros 
     Restriction: fw: 197 micros 
     Interpolation + add: 112.4 micros 
     VCycle: 3812 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 162.6 micros 
     Residual calculation: 168.8 micros 
     Restriction: fw: 200 micros 
     Interpolation + add: 112.4 micros 
     VCycle: 3844 micros 
      
      
      
     ------------------ 
     -- size: 255 
      
     BENCHMARK 
     Jacobi iteration: 3.75 micros 
     Residual calculation: 468.75 micros 
     Restriction: fw: 804.75 micros 
     Interpolation + add: 511.75 micros 
     VCycle: 13672.5 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 582 micros 
     Residual calculation: 668 micros 
     Restriction: fw: 816.5 micros 
     Interpolation + add: 500 micros 
     VCycle: 13672.5 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 550.75 micros 
     Residual calculation: 1453.25 micros 
     Restriction: fw: 43 micros 
     Interpolation + add: 507.75 micros 
     VCycle: 13710 micros 
      
      
      
     ------------------ 
     -- size: 511 
      
     BENCHMARK 
     Jacobi iteration: 6510.67 micros 
     Residual calculation: 1812.33 micros 
     Restriction: fw: 2619.67 micros 
     Interpolation + add: 10.6667 micros 
     VCycle: 53123.3 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 21 micros 
     Residual calculation: 343.667 micros 
     Restriction: fw: 6718.67 micros 
     Interpolation + add: 1010.33 micros 
     VCycle: 40676.7 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 6755.33 micros 
     Residual calculation: 2843.67 micros 
     Restriction: fw: 1739.67 micros 
     Interpolation + add: 2343.67 micros 
     VCycle: 25833.3 micros 
      
      
      
     ------------------ 
     -- size: 1023 
      
     BENCHMARK 
     Jacobi iteration: 12453 micros 
     Residual calculation: 14906 micros 
     Restriction: fw: 4328.5 micros 
     Interpolation + add: 8679.5 micros 
     VCycle: 145155 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 12203 micros 
     Residual calculation: 10297 micros 
     Restriction: fw: 11539 micros 
     Interpolation + add: 9875 micros 
     VCycle: 101640 micros 
      
      
     BENCHMARK 
     Jacobi iteration: 8414.5 micros 
     Residual calculation: 12203 micros 
     Restriction: fw: 9882.5 micros 
     Interpolation + add: 5930 micros 
     VCycle: 135470 micros
    With a 1900XT and Vista32 all the tests ran in about 15 micros. One thing I might suggest is to make sure VPU recover is disabled for the ATI systems. Doubt it will be an issue but it doesn't always like long running shaders.
     
  5. RickCain

    Newcomer

    Joined:
    May 4, 2004
    Messages:
    70
    Likes Received:
    2
    HD2900 running Vista 32-bit

    Code:
    GL multigrid benchmark 0.5 by Peter Thoman 2007
    
    DevIL initialized successfully.
    OpenGL initialized successfully.
    ILUT OpenGL mode set successfully.
    GLEW version 1.3.5 initialized successfully.
    Loaded required OpenGL extensions for GLPixelShader.
    Loaded required OpenGL extensions for GLRenderTexture.
    Loaded required OpenGL extensions for GLFilterStep.
    Initialization complete.
    
    
    
    ------------------
    -- size: 3
    
    BENCHMARK
    Jacobi iteration: 10.8 micros
    Residual calculation: 10.4 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.4 micros
    VCycle: 206 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.3 micros
    Residual calculation: 10.5 micros
    Restriction: fw: 7.7 micros
    Interpolation + add: 10.4 micros
    VCycle: 206 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.3 micros
    Residual calculation: 10.3 micros
    Restriction: fw: 7.7 micros
    Interpolation + add: 10.1 micros
    VCycle: 202 micros
    
    
    
    ------------------
    -- size: 7
    
    BENCHMARK
    Jacobi iteration: 10.5556 micros
    Residual calculation: 10.4444 micros
    Restriction: fw: 7.55556 micros
    Interpolation + add: 10.6667 micros
    VCycle: 326.667 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.3333 micros
    Residual calculation: 10.1111 micros
    Restriction: fw: 7.66667 micros
    Interpolation + add: 10.6667 micros
    VCycle: 323.333 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.7778 micros
    Residual calculation: 10.1111 micros
    Restriction: fw: 7.55556 micros
    Interpolation + add: 10.2222 micros
    VCycle: 343.333 micros
    
    
    
    ------------------
    -- size: 15
    
    BENCHMARK
    Jacobi iteration: 10.75 micros
    Residual calculation: 10.625 micros
    Restriction: fw: 7.875 micros
    Interpolation + add: 10.25 micros
    VCycle: 460 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.5 micros
    Residual calculation: 10.25 micros
    Restriction: fw: 7.625 micros
    Interpolation + add: 10.875 micros
    VCycle: 458.75 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.5 micros
    Residual calculation: 11.75 micros
    Restriction: fw: 7.875 micros
    Interpolation + add: 10.25 micros
    VCycle: 450 micros
    
    
    
    ------------------
    -- size: 31
    
    BENCHMARK
    Jacobi iteration: 10.4286 micros
    Residual calculation: 10.4286 micros
    Restriction: fw: 8.14286 micros
    Interpolation + add: 10.7143 micros
    VCycle: 584.286 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11 micros
    Restriction: fw: 8.57143 micros
    Interpolation + add: 10.5714 micros
    VCycle: 585.714 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.2857 micros
    Residual calculation: 10.7143 micros
    Restriction: fw: 7.71429 micros
    Interpolation + add: 10.2857 micros
    VCycle: 581.429 micros
    
    
    
    ------------------
    -- size: 63
    
    BENCHMARK
    Jacobi iteration: 10.5 micros
    Residual calculation: 10.1667 micros
    Restriction: fw: 7.83333 micros
    Interpolation + add: 10 micros
    VCycle: 695 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.5 micros
    Residual calculation: 10.3333 micros
    Restriction: fw: 7.83333 micros
    Interpolation + add: 10.1667 micros
    VCycle: 698.333 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.1667 micros
    Residual calculation: 10.1667 micros
    Restriction: fw: 7.33333 micros
    Interpolation + add: 10.6667 micros
    VCycle: 681.667 micros
    
    
    
    ------------------
    -- size: 127
    
    BENCHMARK
    Jacobi iteration: 10.4 micros
    Residual calculation: 10 micros
    Restriction: fw: 8.2 micros
    Interpolation + add: 10.2 micros
    VCycle: 880 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.4 micros
    Residual calculation: 11.2 micros
    Restriction: fw: 7.8 micros
    Interpolation + add: 10.2 micros
    VCycle: 838 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.2 micros
    Residual calculation: 10.4 micros
    Restriction: fw: 7.6 micros
    Interpolation + add: 12 micros
    VCycle: 852 micros
    
    
    
    ------------------
    -- size: 255
    
    BENCHMARK
    Jacobi iteration: 10.5 micros
    Residual calculation: 10.5 micros
    Restriction: fw: 7.75 micros
    Interpolation + add: 9.75 micros
    VCycle: 952.5 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.5 micros
    Residual calculation: 10.5 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.25 micros
    VCycle: 975 micros
    
    
    BENCHMARK
    Jacobi iteration: 13.75 micros
    Residual calculation: 13 micros
    Restriction: fw: 8.25 micros
    Interpolation + add: 10.5 micros
    VCycle: 1010 micros
    
    
    
    ------------------
    -- size: 511
    
    BENCHMARK
    Jacobi iteration: 10.3333 micros
    Residual calculation: 10.6667 micros
    Restriction: fw: 8.66667 micros
    Interpolation + add: 10.3333 micros
    VCycle: 1100 micros
    
    
    BENCHMARK
    Jacobi iteration: 10 micros
    Residual calculation: 11 micros
    Restriction: fw: 7.33333 micros
    Interpolation + add: 11 micros
    VCycle: 1086.67 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.3333 micros
    Residual calculation: 11.3333 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.6667 micros
    VCycle: 1053.33 micros
    
    
    
    ------------------
    -- size: 1023
    
    BENCHMARK
    Jacobi iteration: 10.5 micros
    Residual calculation: 11 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10 micros
    VCycle: 1235 micros
    
    
    BENCHMARK
    Jacobi iteration: 12.5 micros
    Residual calculation: 10.5 micros
    Restriction: fw: 9 micros
    Interpolation + add: 11.5 micros
    VCycle: 1280 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.5 micros
    Residual calculation: 10 micros
    Restriction: fw: 8.5 micros
    Interpolation + add: 10.5 micros
    VCycle: 1220 micros
    
    
    
    ------------------
    -- size: 2047
    
    BENCHMARK
    Jacobi iteration: 10 micros
    Residual calculation: 13 micros
    Restriction: fw: 7 micros
    Interpolation + add: 11 micros
    VCycle: 1300 micros
    
    
    BENCHMARK
    Jacobi iteration: 10 micros
    Residual calculation: 10 micros
    Restriction: fw: 7 micros
    Interpolation + add: 11 micros
    VCycle: 1350 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 9 micros
    Restriction: fw: 10 micros
    Interpolation + add: 11 micros
    VCycle: 1360 micros
    
     
  6. Silent_Buddha

    Legend

    Joined:
    Mar 13, 2007
    Messages:
    16,567
    Likes Received:
    5,579
    Yeah, apparently there's a bug in the beta 8.38 release 7 driver that I tested where after running a 3d application the card would remain in HIGH 3D mode and fan would stay at 100% even on the desktop. (LOTS of heat coming out of the rear exhaust.)

    I've since reverted to 8.37 and everything is fine again with it properly reverting to 2D speeds at the desktop.

    I'll run this for you Peter in a moment with both 8.37 and 8.38. Not looking forward to putting 8.38 back on again but Neliz's run appeared to be going maybe it'll complete for me.

    And damn? No 3d clocks in windowed modes? /sigh. I run any game I can in windowed mode when possible. Vanguard, EQ2, Supreme Commander...

    Regards,
    SB
     
  7. BRiT

    BRiT (╯°□°)╯
    Moderator Legend Alpha

    Joined:
    Feb 7, 2002
    Messages:
    13,306
    Likes Received:
    10,024
    Location:
    Cleveland
    Gave it a run on an intermeddiate system while I sort out issues on the new system.

    Vista 64bit run in "Test mode" for experimental drivers
    MSI 8800 GTX HD OC (air cooled model) running stock clocks
    Forceware 158.43 drivers
    AMD X2 4400+ @ 2.5Ghz
    2Gig Mem

    Not sure what to make of these results, as I'm getting some odd behavior with respect to performance in Lost Planet DX10 demo.

    Code:
    GL multigrid benchmark 0.5 by Peter Thoman 2007
    
    DevIL initialized successfully.
    OpenGL initialized successfully.
    ILUT OpenGL mode set successfully.
    GLEW version 1.3.5 initialized successfully.
    Loaded required OpenGL extensions for GLPixelShader.
    Loaded required OpenGL extensions for GLRenderTexture.
    Loaded required OpenGL extensions for GLFilterStep.
    Initialization complete.
    
    
    
    ------------------
    -- size: 3
    
    BENCHMARK
    Jacobi iteration: 4.7 micros
    Residual calculation: 3.1 micros
    Restriction: fw: 3.1 micros
    Interpolation + add: 3.2 micros
    VCycle: 6359 micros
    
    
    BENCHMARK
    Jacobi iteration: 3.2 micros
    Residual calculation: 3.1 micros
    Restriction: fw: 4.7 micros
    Interpolation + add: 3.1 micros
    VCycle: 6469 micros
    
    
    BENCHMARK
    Jacobi iteration: 3.2 micros
    Residual calculation: 3.1 micros
    Restriction: fw: 3.1 micros
    Interpolation + add: 3.1 micros
    VCycle: 6391 micros
    
    
    
    ------------------
    -- size: 7
    
    BENCHMARK
    Jacobi iteration: 5.22222 micros
    Residual calculation: 1.77778 micros
    Restriction: fw: 3.44444 micros
    Interpolation + add: 3.44444 micros
    VCycle: 10555.6 micros
    
    
    BENCHMARK
    Jacobi iteration: 3.44444 micros
    Residual calculation: 3.44444 micros
    Restriction: fw: 3.55556 micros
    Interpolation + add: 3.44444 micros
    VCycle: 10660 micros
    
    
    BENCHMARK
    Jacobi iteration: 3.44444 micros
    Residual calculation: 3.44444 micros
    Restriction: fw: 3.44444 micros
    Interpolation + add: 3.55556 micros
    VCycle: 10972.2 micros
    
    
    
    ------------------
    -- size: 15
    
    BENCHMARK
    Jacobi iteration: 2 micros
    Residual calculation: 3.875 micros
    Restriction: fw: 3.875 micros
    Interpolation + add: 2 micros
    VCycle: 15195 micros
    
    
    BENCHMARK
    Jacobi iteration: 4 micros
    Residual calculation: 3.875 micros
    Restriction: fw: 3.875 micros
    Interpolation + add: 2 micros
    VCycle: 15605 micros
    
    
    BENCHMARK
    Jacobi iteration: 3.875 micros
    Residual calculation: 2 micros
    Restriction: fw: 3.875 micros
    Interpolation + add: 3.875 micros
    VCycle: 15821.3 micros
    
    
    
    ------------------
    -- size: 31
    
    BENCHMARK
    Jacobi iteration: 4.42857 micros
    Residual calculation: 2.28571 micros
    Restriction: fw: 4.42857 micros
    Interpolation + add: 2.14286 micros
    VCycle: 19888.6 micros
    
    
    BENCHMARK
    Jacobi iteration: 2.28571 micros
    Residual calculation: 4.42857 micros
    Restriction: fw: 4.42857 micros
    Interpolation + add: 2.28571 micros
    VCycle: 19420 micros
    
    
    BENCHMARK
    Jacobi iteration: 2.28571 micros
    Residual calculation: 4.42857 micros
    Restriction: fw: 2.28571 micros
    Interpolation + add: 4.42857 micros
    VCycle: 19732.9 micros
    
    
    
    ------------------
    -- size: 63
    
    BENCHMARK
    Jacobi iteration: 2.66667 micros
    Residual calculation: 5.16667 micros
    Restriction: fw: 2.66667 micros
    Interpolation + add: 2.5 micros
    VCycle: 24350 micros
    
    
    BENCHMARK
    Jacobi iteration: 2.5 micros
    Residual calculation: 5.33333 micros
    Restriction: fw: 2.5 micros
    Interpolation + add: 2.66667 micros
    VCycle: 23671.7 micros
    
    
    BENCHMARK
    Jacobi iteration: 2.66667 micros
    Residual calculation: 5.16667 micros
    Restriction: fw: 2.5 micros
    Interpolation + add: 5.33333 micros
    VCycle: 24661.7 micros
    
    
    
    ------------------
    -- size: 127
    
    BENCHMARK
    Jacobi iteration: 3.2 micros
    Residual calculation: 3.2 micros
    Restriction: fw: 6.2 micros
    Interpolation + add: 12.4 micros
    VCycle: 29250 micros
    
    
    BENCHMARK
    Jacobi iteration: 3.2 micros
    Residual calculation: 3 micros
    Restriction: fw: 6.2 micros
    Interpolation + add: 12.6 micros
    VCycle: 28718 micros
    
    
    BENCHMARK
    Jacobi iteration: 3 micros
    Residual calculation: 3.2 micros
    Restriction: fw: 9.4 micros
    Interpolation + add: 9.4 micros
    VCycle: 28406 micros
    
    
    
    ------------------
    -- size: 255
    
    BENCHMARK
    Jacobi iteration: 4 micros
    Residual calculation: 11.75 micros
    Restriction: fw: 46.75 micros
    Interpolation + add: 43 micros
    VCycle: 34882.5 micros
    
    
    BENCHMARK
    Jacobi iteration: 4 micros
    Residual calculation: 15.5 micros
    Restriction: fw: 43 micros
    Interpolation + add: 43 micros
    VCycle: 35235 micros
    
    
    BENCHMARK
    Jacobi iteration: 4 micros
    Residual calculation: 11.75 micros
    Restriction: fw: 46.75 micros
    Interpolation + add: 43 micros
    VCycle: 34220 micros
    
    
    
    ------------------
    -- size: 511
    
    BENCHMARK
    Jacobi iteration: 5.33333 micros
    Residual calculation: 5.33333 micros
    Restriction: fw: 135.333 micros
    Interpolation + add: 192.667 micros
    VCycle: 44896.7 micros
    
    
    BENCHMARK
    Jacobi iteration: 5 micros
    Residual calculation: 5.33333 micros
    Restriction: fw: 135.333 micros
    Interpolation + add: 192.667 micros
    VCycle: 44686.7 micros
    
    
    BENCHMARK
    Jacobi iteration: 5 micros
    Residual calculation: 0 micros
    Restriction: fw: 140.667 micros
    Interpolation + add: 192.667 micros
    VCycle: 45260 micros
    
    
    
    ------------------
    -- size: 1023
    
    BENCHMARK
    Jacobi iteration: 0 micros
    Residual calculation: 0 micros
    Restriction: fw: 8 micros
    Interpolation + add: 578 micros
    VCycle: 82970 micros
    
    
    BENCHMARK
    Jacobi iteration: 0 micros
    Residual calculation: 7.5 micros
    Restriction: fw: 0 micros
    Interpolation + add: 586 micros
    VCycle: 83515 micros
    
    
    BENCHMARK
    Jacobi iteration: 7.5 micros
    Residual calculation: 0 micros
    Restriction: fw: 8 micros
    Interpolation + add: 586 micros
    VCycle: 82810 micros
    
    
    
    ------------------
    -- size: 2047
    
    BENCHMARK
    Jacobi iteration: 0 micros
    Residual calculation: 0 micros
    Restriction: fw: 0 micros
    Interpolation + add: 0 micros
    VCycle: 237190 micros
    
    
    BENCHMARK
    Jacobi iteration: 0 micros
    Residual calculation: 0 micros
    Restriction: fw: 0 micros
    Interpolation + add: 16 micros
    VCycle: 236410 micros
    
    
    BENCHMARK
    Jacobi iteration: 15 micros
    Residual calculation: 0 micros
    Restriction: fw: 0 micros
    Interpolation + add: 0 micros
    VCycle: 238220 micros
    
     
  8. Silent_Buddha

    Legend

    Joined:
    Mar 13, 2007
    Messages:
    16,567
    Likes Received:
    5,579
    Good news and Bad news. 8.37 is obviously bugged, I'm including the scores anyway just for completeness. 8.38 release 7 however your program appears to work fine.

    All runs were done on a core2duo, 4 gigs mem, and Vista 64.

    [Edit - ooops HD 2900 XT]

    8.37.4.3 - default 2d clocks
    Code:
    GL multigrid benchmark 0.5 by Peter Thoman 2007
    
    DevIL initialized successfully.
    OpenGL initialized successfully.
    ILUT OpenGL mode set successfully.
    GLEW version 1.3.5 initialized successfully.
    Loaded required OpenGL extensions for GLPixelShader.
    Loaded required OpenGL extensions for GLRenderTexture.
    Loaded required OpenGL extensions for GLFilterStep.
    Initialization complete.
    
    
    
    ------------------
    -- size: 3
    
    BENCHMARK
    Jacobi iteration: 11.2 micros
    Residual calculation: 10.9 micros
    Restriction: fw: 8.1 micros
    Interpolation + add: 10.8 micros
    VCycle: 212 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.5 micros
    Residual calculation: 10.9 micros
    Restriction: fw: 8.1 micros
    Interpolation + add: 11 micros
    VCycle: 211 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.2 micros
    Residual calculation: 11.3 micros
    Restriction: fw: 8.1 micros
    Interpolation + add: 10.8 micros
    VCycle: 210 micros
    
    
    
    ------------------
    -- size: 7
    
    BENCHMARK
    Jacobi iteration: 11.1111 micros
    Residual calculation: 11.2222 micros
    Restriction: fw: 8.22222 micros
    Interpolation + add: 11.2222 micros
    VCycle: 343.333 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.1111 micros
    Residual calculation: 11 micros
    Restriction: fw: 8.22222 micros
    Interpolation + add: 11 micros
    VCycle: 344.444 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.1111 micros
    Residual calculation: 11.1111 micros
    Restriction: fw: 8.44444 micros
    Interpolation + add: 11 micros
    VCycle: 344.444 micros
    
    
    
    ------------------
    -- size: 15
    
    BENCHMARK
    Jacobi iteration: 11.125 micros
    Residual calculation: 10.875 micros
    Restriction: fw: 8.125 micros
    Interpolation + add: 10.875 micros
    VCycle: 475 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11.125 micros
    Restriction: fw: 8.125 micros
    Interpolation + add: 10.875 micros
    VCycle: 481.25 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.125 micros
    Residual calculation: 11.375 micros
    Restriction: fw: 8.375 micros
    Interpolation + add: 10.875 micros
    VCycle: 475 micros
    
    
    
    ------------------
    -- size: 31
    
    BENCHMARK
    Jacobi iteration: 11.2857 micros
    Residual calculation: 11.1429 micros
    Restriction: fw: 8.28571 micros
    Interpolation + add: 10.7143 micros
    VCycle: 605.714 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.1429 micros
    Residual calculation: 11 micros
    Restriction: fw: 8.57143 micros
    Interpolation + add: 10.8571 micros
    VCycle: 605.714 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11 micros
    Restriction: fw: 8.28571 micros
    Interpolation + add: 10.7143 micros
    VCycle: 618.571 micros
    
    
    
    ------------------
    -- size: 63
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 10.8333 micros
    Restriction: fw: 8.16667 micros
    Interpolation + add: 11.1667 micros
    VCycle: 750 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11 micros
    Restriction: fw: 8 micros
    Interpolation + add: 11 micros
    VCycle: 751.667 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.1667 micros
    Residual calculation: 11 micros
    Restriction: fw: 8.16667 micros
    Interpolation + add: 11 micros
    VCycle: 740 micros
    
    
    
    ------------------
    -- size: 127
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11.2 micros
    Restriction: fw: 8.2 micros
    Interpolation + add: 10.8 micros
    VCycle: 868 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.8 micros
    Residual calculation: 11.2 micros
    Restriction: fw: 8.2 micros
    Interpolation + add: 11 micros
    VCycle: 870 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11.2 micros
    Restriction: fw: 8.4 micros
    Interpolation + add: 10.8 micros
    VCycle: 878 micros
    
    
    
    ------------------
    -- size: 255
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11 micros
    Restriction: fw: 8 micros
    Interpolation + add: 11 micros
    VCycle: 1012.5 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.5 micros
    Residual calculation: 11 micros
    Restriction: fw: 8.25 micros
    Interpolation + add: 11 micros
    VCycle: 1017.5 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.25 micros
    Residual calculation: 10.75 micros
    Restriction: fw: 8.25 micros
    Interpolation + add: 11 micros
    VCycle: 1010 micros
    
    
    
    ------------------
    -- size: 511
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11.3333 micros
    Restriction: fw: 8.66667 micros
    Interpolation + add: 12 micros
    VCycle: 1173.33 micros
    
    
    BENCHMARK
    Jacobi iteration: 12.3333 micros
    Residual calculation: 11.6667 micros
    Restriction: fw: 8 micros
    Interpolation + add: 11 micros
    VCycle: 1143.33 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.3333 micros
    Residual calculation: 11 micros
    Restriction: fw: 8 micros
    Interpolation + add: 11 micros
    VCycle: 1140 micros
    
    
    
    ------------------
    -- size: 1023
    
    BENCHMARK
    Jacobi iteration: 11.5 micros
    Residual calculation: 11 micros
    Restriction: fw: 8.5 micros
    Interpolation + add: 11 micros
    VCycle: 1295 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.5 micros
    Residual calculation: 11.5 micros
    Restriction: fw: 7.5 micros
    Interpolation + add: 11 micros
    VCycle: 1275 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.5 micros
    Residual calculation: 11 micros
    Restriction: fw: 8.5 micros
    Interpolation + add: 11 micros
    VCycle: 1280 micros
    
    
    
    ------------------
    -- size: 2047
    
    BENCHMARK
    Jacobi iteration: 10 micros
    Residual calculation: 12 micros
    Restriction: fw: 9 micros
    Interpolation + add: 11 micros
    VCycle: 1400 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11 micros
    Restriction: fw: 9 micros
    Interpolation + add: 10 micros
    VCycle: 1440 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11 micros
    Restriction: fw: 7 micros
    Interpolation + add: 12 micros
    VCycle: 1400 micros
    
    Beta 8.38 release 7 - default 2d clocks.
    Code:
    GL multigrid benchmark 0.5 by Peter Thoman 2007
    
    DevIL initialized successfully.
    OpenGL initialized successfully.
    ILUT OpenGL mode set successfully.
    GLEW version 1.3.5 initialized successfully.
    Loaded required OpenGL extensions for GLPixelShader.
    Loaded required OpenGL extensions for GLRenderTexture.
    Loaded required OpenGL extensions for GLFilterStep.
    Initialization complete.
    
    
    
    ------------------
    -- size: 3
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 10.9 micros
    Restriction: fw: 8 micros
    Interpolation + add: 11.1 micros
    VCycle: 245 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.1 micros
    Residual calculation: 10.9 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.9 micros
    VCycle: 247 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.1 micros
    Residual calculation: 11 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.7 micros
    VCycle: 246 micros
    
    
    
    ------------------
    -- size: 7
    
    BENCHMARK
    Jacobi iteration: 10.8889 micros
    Residual calculation: 10.6667 micros
    Restriction: fw: 8.11111 micros
    Interpolation + add: 10.8889 micros
    VCycle: 400 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.7778 micros
    Residual calculation: 10.6667 micros
    Restriction: fw: 7.88889 micros
    Interpolation + add: 10.8889 micros
    VCycle: 401.111 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 10.7778 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.8889 micros
    VCycle: 402.222 micros
    
    
    
    ------------------
    -- size: 15
    
    BENCHMARK
    Jacobi iteration: 10.875 micros
    Residual calculation: 10.875 micros
    Restriction: fw: 7.875 micros
    Interpolation + add: 10.875 micros
    VCycle: 552.5 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.75 micros
    Residual calculation: 10.75 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.875 micros
    VCycle: 555 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.75 micros
    Residual calculation: 10.875 micros
    Restriction: fw: 8.125 micros
    Interpolation + add: 10.75 micros
    VCycle: 551.25 micros
    
    
    
    ------------------
    -- size: 31
    
    BENCHMARK
    Jacobi iteration: 10.7143 micros
    Residual calculation: 11.1429 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.8571 micros
    VCycle: 725.714 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.2857 micros
    Residual calculation: 11 micros
    Restriction: fw: 8 micros
    Interpolation + add: 11 micros
    VCycle: 712.857 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11.2857 micros
    Restriction: fw: 8.14286 micros
    Interpolation + add: 11 micros
    VCycle: 714.286 micros
    
    
    
    ------------------
    -- size: 63
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 10.8333 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.6667 micros
    VCycle: 861.667 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.3333 micros
    Residual calculation: 10.8333 micros
    Restriction: fw: 8 micros
    Interpolation + add: 11.1667 micros
    VCycle: 860 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.1667 micros
    Residual calculation: 10.8333 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.8333 micros
    VCycle: 888.333 micros
    
    
    
    ------------------
    -- size: 127
    
    BENCHMARK
    Jacobi iteration: 11.6 micros
    Residual calculation: 11.2 micros
    Restriction: fw: 8.6 micros
    Interpolation + add: 11.2 micros
    VCycle: 1090 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.8 micros
    Residual calculation: 10.6 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.8 micros
    VCycle: 1056 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.6 micros
    Residual calculation: 10.8 micros
    Restriction: fw: 7.6 micros
    Interpolation + add: 10.8 micros
    VCycle: 1046 micros
    
    
    
    ------------------
    -- size: 255
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11 micros
    Restriction: fw: 7.75 micros
    Interpolation + add: 10.75 micros
    VCycle: 4290 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.25 micros
    Residual calculation: 10.75 micros
    Restriction: fw: 8 micros
    Interpolation + add: 10.75 micros
    VCycle: 4590 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.25 micros
    Residual calculation: 10.75 micros
    Restriction: fw: 7.75 micros
    Interpolation + add: 10.75 micros
    VCycle: 4525 micros
    
    
    
    ------------------
    -- size: 511
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 10.6667 micros
    Restriction: fw: 8 micros
    Interpolation + add: 11 micros
    VCycle: 20886.7 micros
    
    
    BENCHMARK
    Jacobi iteration: 16.3333 micros
    Residual calculation: 13.6667 micros
    Restriction: fw: 7.66667 micros
    Interpolation + add: 10.6667 micros
    VCycle: 22576.7 micros
    
    
    BENCHMARK
    Jacobi iteration: 11 micros
    Residual calculation: 11.6667 micros
    Restriction: fw: 11.6667 micros
    Interpolation + add: 12.3333 micros
    VCycle: 22606.7 micros
    
    
    
    ------------------
    -- size: 1023
    
    BENCHMARK
    Jacobi iteration: 15.5 micros
    Residual calculation: 16.5 micros
    Restriction: fw: 11.5 micros
    Interpolation + add: 12 micros
    VCycle: 82325 micros
    
    
    BENCHMARK
    Jacobi iteration: 40 micros
    Residual calculation: 37 micros
    Restriction: fw: 40 micros
    Interpolation + add: 41.5 micros
    VCycle: 89840 micros
    
    
    BENCHMARK
    Jacobi iteration: 40 micros
    Residual calculation: 36.5 micros
    Restriction: fw: 40.5 micros
    Interpolation + add: 41 micros
    VCycle: 89760 micros
    
    
    
    ------------------
    -- size: 2047
    
    BENCHMARK
    Jacobi iteration: 10 micros
    Residual calculation: 11 micros
    Restriction: fw: 8 micros
    Interpolation + add: 11 micros
    VCycle: 1600 micros
    
    
    BENCHMARK
    Jacobi iteration: 6648 micros
    Residual calculation: 6009 micros
    Restriction: fw: 2761 micros
    Interpolation + add: 3047 micros
    VCycle: 50800 micros
    
    
    BENCHMARK
    Jacobi iteration: 6804 micros
    Residual calculation: 5847 micros
    Restriction: fw: 2271 micros
    Interpolation + add: 3460 micros
    VCycle: 50760 micros
    
    Regards,
    SB
     
  9. SugarCoat

    Veteran

    Joined:
    Jul 17, 2005
    Messages:
    2,091
    Likes Received:
    52
    Location:
    State of Illusionism
    Windows Server 2003 64Bit (build 3790, Service Pack 2)
    Intel(R) Core(TM)2 CPU 6300 @ 2.407GHz
    NVIDIA GeForce 8800 GTX 6.14.11.5819 768Mb

    Code:
    GL multigrid benchmark 0.5 by Peter Thoman 2007
    
    DevIL initialized successfully.
    OpenGL initialized successfully.
    ILUT OpenGL mode set successfully.
    GLEW version 1.3.5 initialized successfully.
    Loaded required OpenGL extensions for GLPixelShader.
    Loaded required OpenGL extensions for GLRenderTexture.
    Loaded required OpenGL extensions for GLFilterStep.
    Initialization complete.
    
    
    
    ------------------
    -- size: 3
    
    BENCHMARK
    Jacobi iteration: 3.1 micros
    Residual calculation: 3.1 micros
    Restriction: fw: 1.6 micros
    Interpolation + add: 1.5 micros
    VCycle: 547 micros
    
    
    BENCHMARK
    Jacobi iteration: 1.5 micros
    Residual calculation: 3.2 micros
    Restriction: fw: 1.5 micros
    Interpolation + add: 3.2 micros
    VCycle: 546 micros
    
    
    BENCHMARK
    Jacobi iteration: 3.2 micros
    Residual calculation: 3.1 micros
    Restriction: fw: 1.5 micros
    Interpolation + add: 3.2 micros
    VCycle: 547 micros
    
    
    
    ------------------
    -- size: 7
    
    BENCHMARK
    Jacobi iteration: 3.44444 micros
    Residual calculation: 1.77778 micros
    Restriction: fw: 3.44444 micros
    Interpolation + add: 1.77778 micros
    VCycle: 920 micros
    
    
    BENCHMARK
    Jacobi iteration: 3.44444 micros
    Residual calculation: 1.77778 micros
    Restriction: fw: 1.66667 micros
    Interpolation + add: 3.55556 micros
    VCycle: 920 micros
    
    
    BENCHMARK
    Jacobi iteration: 3.44444 micros
    Residual calculation: 1.66667 micros
    Restriction: fw: 3.55556 micros
    Interpolation + add: 1.66667 micros
    VCycle: 937.778 micros
    
    
    
    ------------------
    -- size: 15
    
    BENCHMARK
    Jacobi iteration: 3.875 micros
    Residual calculation: 1.875 micros
    Restriction: fw: 2 micros
    Interpolation + add: 2 micros
    VCycle: 1288.75 micros
    
    
    BENCHMARK
    Jacobi iteration: 1.875 micros
    Residual calculation: 4 micros
    Restriction: fw: 1.875 micros
    Interpolation + add: 2 micros
    VCycle: 1288.75 micros
    
    
    BENCHMARK
    Jacobi iteration: 1.875 micros
    Residual calculation: 2 micros
    Restriction: fw: 2 micros
    Interpolation + add: 1.875 micros
    VCycle: 1308.75 micros
    
    
    
    ------------------
    -- size: 31
    
    BENCHMARK
    Jacobi iteration: 2.28571 micros
    Residual calculation: 4.42857 micros
    Restriction: fw: 2.28571 micros
    Interpolation + add: 2.14286 micros
    VCycle: 1652.86 micros
    
    
    BENCHMARK
    Jacobi iteration: 2.28571 micros
    Residual calculation: 4.42857 micros
    Restriction: fw: 2.28571 micros
    Interpolation + add: 2.28571 micros
    VCycle: 1672.86 micros
    
    
    BENCHMARK
    Jacobi iteration: 2.28571 micros
    Residual calculation: 2.28571 micros
    Restriction: fw: 2.14286 micros
    Interpolation + add: 2.28571 micros
    VCycle: 1674.29 micros
    
    
    
    ------------------
    -- size: 63
    
    BENCHMARK
    Jacobi iteration: 5.33333 micros
    Residual calculation: 2.5 micros
    Restriction: fw: 2.66667 micros
    Interpolation + add: 2.66667 micros
    VCycle: 2056.67 micros
    
    
    BENCHMARK
    Jacobi iteration: 2.5 micros
    Residual calculation: 5.33333 micros
    Restriction: fw: 2.5 micros
    Interpolation + add: 2.66667 micros
    VCycle: 2056.67 micros
    
    
    BENCHMARK
    Jacobi iteration: 2.5 micros
    Residual calculation: 2.66667 micros
    Restriction: fw: 2.66667 micros
    Interpolation + add: 2.5 micros
    VCycle: 2058.33 micros
    
    
    
    ------------------
    -- size: 127
    
    BENCHMARK
    Jacobi iteration: 12.4 micros
    Residual calculation: 9.4 micros
    Restriction: fw: 12.6 micros
    Interpolation + add: 6.2 micros
    VCycle: 2468 micros
    
    
    BENCHMARK
    Jacobi iteration: 9.4 micros
    Residual calculation: 12.6 micros
    Restriction: fw: 9.2 micros
    Interpolation + add: 6.4 micros
    VCycle: 2468 micros
    
    
    BENCHMARK
    Jacobi iteration: 9.4 micros
    Residual calculation: 12.4 micros
    Restriction: fw: 9.4 micros
    Interpolation + add: 6.2 micros
    VCycle: 2470 micros
    
    
    
    ------------------
    -- size: 255
    
    BENCHMARK
    Jacobi iteration: 46.75 micros
    Residual calculation: 43 micros
    Restriction: fw: 43 micros
    Interpolation + add: 31.25 micros
    VCycle: 3085 micros
    
    
    BENCHMARK
    Jacobi iteration: 46.75 micros
    Residual calculation: 47 micros
    Restriction: fw: 39 micros
    Interpolation + add: 31.25 micros
    VCycle: 3047.5 micros
    
    
    BENCHMARK
    Jacobi iteration: 47 micros
    Residual calculation: 43 micros
    Restriction: fw: 42.75 micros
    Interpolation + add: 31.25 micros
    VCycle: 3047.5 micros
    
    
    
    ------------------
    -- size: 511
    
    BENCHMARK
    Jacobi iteration: 182.333 micros
    Residual calculation: 182.333 micros
    Restriction: fw: 182.333 micros
    Interpolation + add: 114.333 micros
    VCycle: 4533.33 micros
    
    
    BENCHMARK
    Jacobi iteration: 187.333 micros
    Residual calculation: 182.333 micros
    Restriction: fw: 177.333 micros
    Interpolation + add: 119.667 micros
    VCycle: 4530 micros
    
    
    BENCHMARK
    Jacobi iteration: 187.333 micros
    Residual calculation: 177 micros
    Restriction: fw: 182.333 micros
    Interpolation + add: 120 micros
    VCycle: 4476.67 micros
    
    
    
    ------------------
    -- size: 1023
    
    BENCHMARK
    Jacobi iteration: 750 micros
    Residual calculation: 726.5 micros
    Restriction: fw: 797 micros
    Interpolation + add: 523.5 micros
    VCycle: 15390 micros
    
    
    BENCHMARK
    Jacobi iteration: 750 micros
    Residual calculation: 719 micros
    Restriction: fw: 804.5 micros
    Interpolation + add: 515.5 micros
    VCycle: 15390 micros
    
    
    BENCHMARK
    Jacobi iteration: 742 micros
    Residual calculation: 734.5 micros
    Restriction: fw: 789 micros
    Interpolation + add: 523.5 micros
    VCycle: 15390 micros
    
    
    
    ------------------
    -- size: 2047
    
    BENCHMARK
    Jacobi iteration: 437 micros
    Residual calculation: 2406 micros
    Restriction: fw: 3375 micros
    Interpolation + add: 2704 micros
    VCycle: 92650 micros
    
    
    BENCHMARK
    Jacobi iteration: 422 micros
    Residual calculation: 2391 micros
    Restriction: fw: 3422 micros
    Interpolation + add: 2734 micros
    VCycle: 92660 micros
    
    
    BENCHMARK
    Jacobi iteration: 422 micros
    Residual calculation: 2422 micros
    Restriction: fw: 3390 micros
    Interpolation + add: 2781 micros
    VCycle: 92500 micros
    

    They havent run in 3D clocks without fullscreen since they started doing dynamic clocks. Its something that has annoyed me since the introduction of the X1800 and i've commented on a couple times. The fact that the card stays at 2D clocks even with a windowed 3D app strikes me as lazy.
     
    #49 SugarCoat, May 30, 2007
    Last edited by a moderator: May 30, 2007
  10. Silent_Buddha

    Legend

    Joined:
    Mar 13, 2007
    Messages:
    16,567
    Likes Received:
    5,579
    Yeah, not to happy about that. Especially now with Vista and the fact that a window that spans multiple displays remains fully 3d accelerated.

    Regards,
    SB
     
  11. SugarCoat

    Veteran

    Joined:
    Jul 17, 2005
    Messages:
    2,091
    Likes Received:
    52
    Location:
    State of Illusionism
    i use to run a few instances of games windowed when i first noticed it by running rivatuner in the background monitoring the hardware. I didnt even think to check to be honest, you'd think that they wouldnt be so lazy not to fix/include something like that, yet here we are, coming on the 2 year mark since the introduction of the R520...apparently nobody cares. At least not enough to make a fuss and get it fixed in a release or two. As it stands its problably something we wont see resolved until the R7xx refresh cards, it will be a small, size 2 font, notation in the driver release notes. :lol:
     
  12. AlexV

    AlexV Heteroscedasticitate
    Moderator Veteran

    Joined:
    Mar 15, 2005
    Messages:
    2,528
    Likes Received:
    107
    It`s still screwy for 2047...too fast. Otherwise everything seems in line with what(I guess) should be happening. Interesting.
     
  13. banksie

    Newcomer

    Joined:
    Jun 9, 2003
    Messages:
    213
    Likes Received:
    4
    Location:
    Wellington, New Zealand
    Here ya go :-

    Windows XP Pro
    AMD A64 3000+ (Socket 754)
    Radeon HD 2900 XT
    Catalyst 8.37.4.3-070510a-047096E-ATI

    All stock clocks.

    Code:
    GL multigrid benchmark 0.5 by Peter Thoman 2007
    
    DevIL initialized successfully.
    OpenGL initialized successfully.
    ILUT OpenGL mode set successfully.
    GLEW version 1.3.5 initialized successfully.
    Loaded required OpenGL extensions for GLPixelShader.
    Loaded required OpenGL extensions for GLRenderTexture.
    Loaded required OpenGL extensions for GLFilterStep.
    Initialization complete.
    
    
    
    ------------------
    -- size: 3
    
    BENCHMARK
    Jacobi iteration: 14 micros
    Residual calculation: 14.1 micros
    Restriction: fw: 9.4 micros
    Interpolation + add: 14 micros
    VCycle: 282 micros
    
    
    BENCHMARK
    Jacobi iteration: 14 micros
    Residual calculation: 14.1 micros
    Restriction: fw: 9.4 micros
    Interpolation + add: 14 micros
    VCycle: 282 micros
    
    
    BENCHMARK
    Jacobi iteration: 14.1 micros
    Residual calculation: 14 micros
    Restriction: fw: 11 micros
    Interpolation + add: 14 micros
    VCycle: 282 micros
    
    
    
    ------------------
    -- size: 7
    
    BENCHMARK
    Jacobi iteration: 13.8889 micros
    Residual calculation: 13.8889 micros
    Restriction: fw: 10.4444 micros
    Interpolation + add: 15.5556 micros
    VCycle: 434.444 micros
    
    
    BENCHMARK
    Jacobi iteration: 13.8889 micros
    Residual calculation: 15.5556 micros
    Restriction: fw: 10.4444 micros
    Interpolation + add: 13.8889 micros
    VCycle: 451.111 micros
    
    
    BENCHMARK
    Jacobi iteration: 15.6667 micros
    Residual calculation: 13.8889 micros
    Restriction: fw: 10.3333 micros
    Interpolation + add: 13.8889 micros
    VCycle: 468.889 micros
    
    
    
    ------------------
    -- size: 15
    
    BENCHMARK
    Jacobi iteration: 13.75 micros
    Residual calculation: 13.625 micros
    Restriction: fw: 11.75 micros
    Interpolation + add: 13.625 micros
    VCycle: 606.25 micros
    
    
    BENCHMARK
    Jacobi iteration: 13.75 micros
    Residual calculation: 13.625 micros
    Restriction: fw: 11.75 micros
    Interpolation + add: 13.625 micros
    VCycle: 606.25 micros
    
    
    BENCHMARK
    Jacobi iteration: 13.75 micros
    Residual calculation: 15.625 micros
    Restriction: fw: 9.75 micros
    Interpolation + add: 13.625 micros
    VCycle: 606.25 micros
    
    
    
    ------------------
    -- size: 31
    
    BENCHMARK
    Jacobi iteration: 15.5714 micros
    Residual calculation: 13.4286 micros
    Restriction: fw: 11.1429 micros
    Interpolation + add: 13.4286 micros
    VCycle: 781.429 micros
    
    
    BENCHMARK
    Jacobi iteration: 13.4286 micros
    Residual calculation: 15.5714 micros
    Restriction: fw: 11.2857 micros
    Interpolation + add: 13.2857 micros
    VCycle: 781.429 micros
    
    
    BENCHMARK
    Jacobi iteration: 15.5714 micros
    Residual calculation: 13.4286 micros
    Restriction: fw: 11.1429 micros
    Interpolation + add: 13.4286 micros
    VCycle: 781.429 micros
    
    
    
    ------------------
    -- size: 63
    
    BENCHMARK
    Jacobi iteration: 13 micros
    Residual calculation: 15.6667 micros
    Restriction: fw: 10.5 micros
    Interpolation + add: 13 micros
    VCycle: 963.333 micros
    
    
    BENCHMARK
    Jacobi iteration: 13 micros
    Residual calculation: 13 micros
    Restriction: fw: 10.5 micros
    Interpolation + add: 15.6667 micros
    VCycle: 988.333 micros
    
    
    BENCHMARK
    Jacobi iteration: 15.6667 micros
    Residual calculation: 15.6667 micros
    Restriction: fw: 10.3333 micros
    Interpolation + add: 15.6667 micros
    VCycle: 963.333 micros
    
    
    
    ------------------
    -- size: 127
    
    BENCHMARK
    Jacobi iteration: 12.6 micros
    Residual calculation: 15.6 micros
    Restriction: fw: 9.4 micros
    Interpolation + add: 15.6 micros
    VCycle: 1124 micros
    
    
    BENCHMARK
    Jacobi iteration: 15.8 micros
    Residual calculation: 12.4 micros
    Restriction: fw: 12.6 micros
    Interpolation + add: 12.4 micros
    VCycle: 1156 micros
    
    
    BENCHMARK
    Jacobi iteration: 15.6 micros
    Residual calculation: 12.6 micros
    Restriction: fw: 12.4 micros
    Interpolation + add: 12.6 micros
    VCycle: 1124 micros
    
    
    
    ------------------
    -- size: 255
    
    BENCHMARK
    Jacobi iteration: 11.75 micros
    Residual calculation: 15.5 micros
    Restriction: fw: 8 micros
    Interpolation + add: 15.5 micros
    VCycle: 1290 micros
    
    
    BENCHMARK
    Jacobi iteration: 15.75 micros
    Residual calculation: 15.5 micros
    Restriction: fw: 7.75 micros
    Interpolation + add: 15.75 micros
    VCycle: 1290 micros
    
    
    BENCHMARK
    Jacobi iteration: 11.5 micros
    Residual calculation: 15.75 micros
    Restriction: fw: 7.75 micros
    Interpolation + add: 15.75 micros
    VCycle: 1327.5 micros
    
    
    
    ------------------
    -- size: 511
    
    BENCHMARK
    Jacobi iteration: 10.6667 micros
    Residual calculation: 15.6667 micros
    Restriction: fw: 10.3333 micros
    Interpolation + add: 15.6667 micros
    VCycle: 1456.67 micros
    
    
    BENCHMARK
    Jacobi iteration: 15.6667 micros
    Residual calculation: 15.6667 micros
    Restriction: fw: 10.3333 micros
    Interpolation + add: 15.6667 micros
    VCycle: 1460 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.6667 micros
    Residual calculation: 15.6667 micros
    Restriction: fw: 10.3333 micros
    Interpolation + add: 15.6667 micros
    VCycle: 1456.67 micros
    
    
    
    ------------------
    -- size: 1023
    
    BENCHMARK
    Jacobi iteration: 8 micros
    Residual calculation: 15.5 micros
    Restriction: fw: 16 micros
    Interpolation + add: 7.5 micros
    VCycle: 1640 micros
    
    
    BENCHMARK
    Jacobi iteration: 15.5 micros
    Residual calculation: 15.5 micros
    Restriction: fw: 8 micros
    Interpolation + add: 15.5 micros
    VCycle: 1640 micros
    
    
    BENCHMARK
    Jacobi iteration: 15.5 micros
    Residual calculation: 15.5 micros
    Restriction: fw: 8 micros
    Interpolation + add: 15.5 micros
    VCycle: 1720 micros
    
    
    
    ------------------
    -- size: 2047
    
    BENCHMARK
    Jacobi iteration: 16 micros
    Residual calculation: 16 micros
    Restriction: fw: 0 micros
    Interpolation + add: 15 micros
    VCycle: 1880 micros
    
    
    BENCHMARK
    Jacobi iteration: 15 micros
    Residual calculation: 16 micros
    Restriction: fw: 0 micros
    Interpolation + add: 15 micros
    VCycle: 1880 micros
    
    
    BENCHMARK
    Jacobi iteration: 16 micros
    Residual calculation: 15 micros
    Restriction: fw: 16 micros
    Interpolation + add: 16 micros
    VCycle: 1870 micros
    
     
  14. AlexV

    AlexV Heteroscedasticitate
    Moderator Veteran

    Joined:
    Mar 15, 2005
    Messages:
    2,528
    Likes Received:
    107
    It seems the 8.37 drivers are the culprits, with the 8.38s doing what should be done...so ATi, now you actually have a valid reason for releasing new Catalysts...this man's master thesis must have valid benchmarks backing it up!:D
     
  15. SuperBORG

    Newcomer

    Joined:
    Jun 26, 2004
    Messages:
    30
    Likes Received:
    0
    HD 2900XT 512MB Stock Clocks
    Catalyst 8.39
    Vista 64bit
    4GB RAM

    Code:
    GL multigrid benchmark 0.5 by Peter Thoman 2007
    
    DevIL initialized successfully.
    OpenGL initialized successfully.
    ILUT OpenGL mode set successfully.
    GLEW version 1.3.5 initialized successfully.
    Loaded required OpenGL extensions for GLPixelShader.
    Loaded required OpenGL extensions for GLRenderTexture.
    Loaded required OpenGL extensions for GLFilterStep.
    Initialization complete.
    
    
    
    ------------------
    -- size: 3
    
    BENCHMARK
    Jacobi iteration: 7.8 micros
    Residual calculation: 7.8 micros
    Restriction: fw: 6.2 micros
    Interpolation + add: 7.8 micros
    VCycle: 827 micros
    
    
    BENCHMARK
    Jacobi iteration: 7.8 micros
    Residual calculation: 6.2 micros
    Restriction: fw: 7.8 micros
    Interpolation + add: 6.2 micros
    VCycle: 827 micros
    
    
    BENCHMARK
    Jacobi iteration: 7.8 micros
    Residual calculation: 6.2 micros
    Restriction: fw: 6.2 micros
    Interpolation + add: 7.8 micros
    VCycle: 827 micros
    
    
    
    ------------------
    -- size: 7
    
    BENCHMARK
    Jacobi iteration: 7 micros
    Residual calculation: 6.88889 micros
    Restriction: fw: 7 micros
    Interpolation + add: 6.88889 micros
    VCycle: 1560 micros
    
    
    BENCHMARK
    Jacobi iteration: 7 micros
    Residual calculation: 6.88889 micros
    Restriction: fw: 7 micros
    Interpolation + add: 6.88889 micros
    VCycle: 1542.22 micros
    
    
    BENCHMARK
    Jacobi iteration: 6.88889 micros
    Residual calculation: 7 micros
    Restriction: fw: 6.88889 micros
    Interpolation + add: 6.88889 micros
    VCycle: 1560 micros
    
    
    
    ------------------
    -- size: 15
    
    BENCHMARK
    Jacobi iteration: 5.75 micros
    Residual calculation: 7.875 micros
    Restriction: fw: 5.875 micros
    Interpolation + add: 7.75 micros
    VCycle: 2301.25 micros
    
    
    BENCHMARK
    Jacobi iteration: 7.75 micros
    Residual calculation: 5.875 micros
    Restriction: fw: 7.75 micros
    Interpolation + add: 5.875 micros
    VCycle: 2320 micros
    
    
    BENCHMARK
    Jacobi iteration: 5.875 micros
    Residual calculation: 7.75 micros
    Restriction: fw: 5.875 micros
    Interpolation + add: 7.75 micros
    VCycle: 2301.25 micros
    
    
    
    ------------------
    -- size: 31
    
    BENCHMARK
    Jacobi iteration: 6.71429 micros
    Residual calculation: 8.85714 micros
    Restriction: fw: 4.57143 micros
    Interpolation + add: 8.85714 micros
    VCycle: 3052.86 micros
    
    
    BENCHMARK
    Jacobi iteration: 6.57143 micros
    Residual calculation: 6.71429 micros
    Restriction: fw: 6.71429 micros
    Interpolation + add: 6.71429 micros
    VCycle: 3075.71 micros
    
    
    BENCHMARK
    Jacobi iteration: 9 micros
    Residual calculation: 6.71429 micros
    Restriction: fw: 6.57143 micros
    Interpolation + add: 6.71429 micros
    VCycle: 3075.71 micros
    
    
    
    ------------------
    -- size: 63
    
    BENCHMARK
    Jacobi iteration: 7.66667 micros
    Residual calculation: 7.83333 micros
    Restriction: fw: 7.83333 micros
    Interpolation + add: 5.16667 micros
    VCycle: 3821.67 micros
    
    
    BENCHMARK
    Jacobi iteration: 5.16667 micros
    Residual calculation: 7.83333 micros
    Restriction: fw: 7.83333 micros
    Interpolation + add: 5.16667 micros
    VCycle: 3821.67 micros
    
    
    BENCHMARK
    Jacobi iteration: 5.16667 micros
    Residual calculation: 7.83333 micros
    Restriction: fw: 7.83333 micros
    Interpolation + add: 5.16667 micros
    VCycle: 3821.67 micros
    
    
    
    ------------------
    -- size: 127
    
    BENCHMARK
    Jacobi iteration: 6.2 micros
    Residual calculation: 9.4 micros
    Restriction: fw: 6.2 micros
    Interpolation + add: 9.4 micros
    VCycle: 4616 micros
    
    
    BENCHMARK
    Jacobi iteration: 6.2 micros
    Residual calculation: 6.4 micros
    Restriction: fw: 6.2 micros
    Interpolation + add: 9.4 micros
    VCycle: 4554 micros
    
    
    BENCHMARK
    Jacobi iteration: 6.2 micros
    Residual calculation: 6.2 micros
    Restriction: fw: 6.4 micros
    Interpolation + add: 9.2 micros
    VCycle: 4556 micros
    
    
    
    ------------------
    -- size: 255
    
    BENCHMARK
    Jacobi iteration: 7.75 micros
    Residual calculation: 4 micros
    Restriction: fw: 7.75 micros
    Interpolation + add: 7.75 micros
    VCycle: 8815 micros
    
    
    BENCHMARK
    Jacobi iteration: 7.75 micros
    Residual calculation: 4 micros
    Restriction: fw: 7.75 micros
    Interpolation + add: 7.75 micros
    VCycle: 8892.5 micros
    
    
    BENCHMARK
    Jacobi iteration: 8 micros
    Residual calculation: 7.75 micros
    Restriction: fw: 7.75 micros
    Interpolation + add: 7.75 micros
    VCycle: 8852.5 micros
    
    
    
    ------------------
    -- size: 511
    
    BENCHMARK
    Jacobi iteration: 5.33333 micros
    Residual calculation: 5 micros
    Restriction: fw: 10.3333 micros
    Interpolation + add: 5.33333 micros
    VCycle: 30836.7 micros
    
    
    BENCHMARK
    Jacobi iteration: 5.33333 micros
    Residual calculation: 10.3333 micros
    Restriction: fw: 5 micros
    Interpolation + add: 10.6667 micros
    VCycle: 32913.3 micros
    
    
    BENCHMARK
    Jacobi iteration: 10.3333 micros
    Residual calculation: 5.33333 micros
    Restriction: fw: 5 micros
    Interpolation + add: 15.6667 micros
    VCycle: 32863.3 micros
    
    
    
    ------------------
    -- size: 1023
    
    BENCHMARK
    Jacobi iteration: 8 micros
    Residual calculation: 7.5 micros
    Restriction: fw: 8 micros
    Interpolation + add: 8 micros
    VCycle: 119340 micros
    
    
    BENCHMARK
    Jacobi iteration: 62.5 micros
    Residual calculation: 54.5 micros
    Restriction: fw: 62.5 micros
    Interpolation + add: 54.5 micros
    VCycle: 129405 micros
    
    
    BENCHMARK
    Jacobi iteration: 62 micros
    Residual calculation: 55 micros
    Restriction: fw: 62 micros
    Interpolation + add: 55 micros
    VCycle: 129480 micros
    
    
    
    ------------------
    -- size: 2047
    
    BENCHMARK
    Jacobi iteration: 0 micros
    Residual calculation: 16 micros
    Restriction: fw: 0 micros
    Interpolation + add: 15 micros
    VCycle: 7340 micros
    
    
    BENCHMARK
    Jacobi iteration: 5382 micros
    Residual calculation: 4836 micros
    Restriction: fw: 2745 micros
    Interpolation + add: 203 micros
    VCycle: 68800 micros
    
    
    BENCHMARK
    Jacobi iteration: 5507 micros
    Residual calculation: 4898 micros
    Restriction: fw: 2106 micros
    Interpolation + add: 203 micros
    VCycle: 68800 micros
    
     
    #55 SuperBORG, May 30, 2007
    Last edited by a moderator: May 30, 2007
  16. neliz

    neliz GIGABYTE Man
    Veteran

    Joined:
    Mar 30, 2005
    Messages:
    4,904
    Likes Received:
    23
    Location:
    In the know
    borg, remove the spaces in the [ code ]
     
  17. Blacklash

    Newcomer

    Joined:
    Feb 26, 2004
    Messages:
    219
    Likes Received:
    3
    #57 Blacklash, May 30, 2007
    Last edited by a moderator: May 30, 2007
  18. PeterT

    Regular

    Joined:
    May 14, 2002
    Messages:
    702
    Likes Received:
    14
    Location:
    Austria
    Thanks everybody for all the results, especially Silent_Buddha for running 2 different driver versions. From what I can see, the post-8.37 results are more plausible at least than those at 8.37, but still very strange:
    - The V-cycle times (those are the most important ones by far) seem to be in line with expectations up to 1023^2, if a bit disappointing compared to G80 results.
    - The other 4 values are all still far too good to be true.
    - at 2047^2, all bets are off. I have no idea what's going on there.

    I will try to release a functionality testing app later today to see which (if any) driver versions actually perform the computations correctly.
     
  19. Silent_Buddha

    Legend

    Joined:
    Mar 13, 2007
    Messages:
    16,567
    Likes Received:
    5,579
    PeterT - I'm not sure what's happening however, when actually running mgbench under 2047^2 it took a little over twice as long to finish as it did for 1023^2. I'm not sure why it then recorded cycle times that were lower than 1023^2.

    Regards,
    SB
     
  20. Sxotty

    Veteran

    Joined:
    Dec 11, 2002
    Messages:
    4,970
    Likes Received:
    370
    Location:
    PA USA
    Well my x1900 AIW just crashed and it is at stock clocks... ah well
     
Loading...

Share This Page

  • About Us

    Beyond3D has been around for over a decade and prides itself on being the best place on the web for in-depth, technically-driven discussion and analysis of 3D graphics hardware. If you love pixels and transistors, you've come to the right place!

    Beyond3D is proudly published by GPU Tools Ltd.
Loading...