Closed Thread
Results 1 to 2 of 2

Thread: 3ds max viewport performance test - with 9000 animate particle

  1. #1
    Junior Member kianoosh's Avatar
    Join Date
    Sep 2010
    Posts
    5

    Name
    Kianoush Nouri
    Forum Username
    kianoosh

    Iran, Islamic Republic of

    Default 3ds max viewport performance test - with 9000 animate particle

    please test this scene with your gpu . i am interesting to test by quadro 4000 and up , ati v7900 and up ... and gtx 6xx

    i made 9000 particle - instead object is teapot with 4 segment ( 1024 polygon per each teapot ) , total : 9,216,000 poly
    test on Opengl and DirectX nitrous please. quadro and ATi firepro card have better performance in OpenGL
    just click play animation .

    http://www.sendspace.com/file/rz1o7j

    thanks ....

    my fps : below 2 - on DirectX nitrous
    my spec : 3ds max 2012 64bit & win 7
    ram : 16gb 1600mhz
    cpu : 2600k @4.4 ghz
    VGA : gtx 470 zotac
    vga driver : 310.70 WHQL


    !!!!! i tested on 3ds max 2013 64bit - my fps is 45-50 on DirectX nitrous !!!!

  2. #2
    Senior Member spacefrog's Avatar
    Join Date
    Jan 2006
    Location
    Vienna
    Posts
    250

    Name
    Josef Wienerroither
    Forum Username
    spacefrog

    Austria

    Default Re: 3ds max viewport performance test - with 9000 animate particle

    Before this evolves in an unrepresentative performance comparision a few remarks. I changed all the mentioned settings in the atttached scene...

    * The scene has the vray renderer as active renderer. Not everyone has that renderer installed. This will not affect the result, but its unnecessary, so i removed it

    * Displaying geometry stats in the viewport affects playback performance (slowdown because of modifier stack reevaluation), so those are better disabled. This is not very relevant in this example, but nevertheless should always be done in benchmark scenes

    * The scene has "Realtime Playback enabled", meaning that 3ds Max will try to skip frames to achieve the set goal of NTSC playback ( 30 FPS) . This makes comparisons unrelieable. So first step is to turn it off. Should always be done in VP benchmark scenes

    * The scene uses "Nitrous - Realistic Mode". This is not really compareable to any of the existing DirectX/OpenGL modes. I would recommend to compare "Nitrous - Shaded Mode" with DirectX/OpenGL modes having Hardware shading disabled. I did so in the attached scene

    * The scene has obviously been set up in DirectX or OpenGL mode with viewport degratation enabled (stored in the scene file). Nitrous doesn't have this feature yet, thus you are comparing viewport performance with degratation enabled ( DirectX/OpenGL) with performance having it disabled/Nonexistant (nitrous ). You can see this on slower systems when the "Progressive Display" button at the bottom of the screen goes cyan/green. You only can disable that by pressing the shortcut when in Nitrous mode, because the UI is completly missing than. So i disabled the "Viewport degratation" feature in the attached scene

    * You have your Particle display set to "Bounding Box" which does'nt work in Nitrous mode, Nitrous always displays geometry. So this would compare geometry particles (Nitrous) with bounding box particles ( DirectX/OpenGL). I changed that in the attached scene.

    * finally the Max viewport FPS stats are not really reliable. It would be better measuring the timeline play through time in seconds and divide that by the number of frames in the timeline -> reliable REAL FPS... I have a great script that does this automatically, but for this case it's probably overkill. An alternative would be using tools like Fraps, which display far more correct FPS times

    So after above mentioned changes, my results are:

    Hardware Info:
    W7 x64, Quad Xeon 3350/3.4GHz, 8GB, Geforce GTX 275/1.8GB

    Max 2013 x64 / PU06:
    First value is when particles are displayed as geometry, second with bounding box particles

    Nitrous: ~47fps | ~47fps ( BB mode not working=same as geometry mode)

    DirectX: ~20fps | ~12 fps

    OpenGL: unusable* | ~68 fps:

    *unusable: starts at 3FPS, drops immediatly to below 1fps. This is mainly because of the missing geometry caching that DirectX and Nitrous modes are using


    Please use the modified scene attached below!
    Attached Files
    Last edited by spacefrog; January 2nd, 2013 at 03:30 AM.

Closed Thread

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

     

Similar Threads

  1. Graphic Card for better viewport performance
    By remeusromulus in forum Hardware and Technical Discussions
    Replies: 1
    Last Post: September 17th, 2012, 12:33 AM
  2. Grass Test using particle flow and MR
    By jinsley in forum Vegetation
    Replies: 90
    Last Post: May 25th, 2011, 02:34 AM
  3. Viewport performance
    By joemillson in forum Hardware and Technical Discussions
    Replies: 3
    Last Post: May 6th, 2010, 07:44 AM
  4. Replies: 7
    Last Post: October 21st, 2009, 04:56 AM
  5. optimise viewport performance + workflow
    By max.m in forum Hardware and Technical Discussions
    Replies: 5
    Last Post: September 24th, 2008, 01:07 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts