simulate more than 2000 and up to 65000 floating objects

If I want to simulate more than 2000 and up to 65000 floating objects, what should I do?
The FAQ shows that I have to use GenCase_Mk_Word.exe and DualSPHysics compiled with an special option. Does it means that I need only to activate the code line #define CODE_SIZE4 in Types.h? After activating it, what should I do?

Comments

  • After activating it, what should I do?
    Compile AGAIN and run the case generated with GenCase_Mk_Word.exe with the version of DualSPHysics you have compiled...

    We assumed that the explanation was clear, if not we can try to complete it
  • Thanks, Alex. I have followed the steps of section 7.3 (Compile instructions for Microsoft Windows with Cmake), and generated a series files as follows:
    CMakeFiles (folder)
    ALL_BUILD
    cmake_install.cmake
    CMakeCache.txt
    DualSPHysics
    DualSPHysics4.2
    DualSPHysics4.2.vcxproj
    DualSPHysics4.2CPU
    DualSPHysics4.2CPU.vcxproj
    INSTALL
    INSTALL.vcxproj
    ZERO_CHECK
    ZERO_CHECK.vcxproj

    But I didn't get the DualSPHysics4.2.exe or DualSPHysics4.2CPU.exe files, what should I do in the next step. I have never did something like this before. Thank you for your replying!
  • Have you tried to compile using VS2015 and the provided .sln?
    Regards
  • Where do I define Release or ReleaseCPU?
  • I read the guide and looked up for an answer on this discussion board. As far as I understand, I do not need CUDA installed if I'm running on CPU. Is that correct?
    If it is, where do I define which version I'm using?
  • If you are running on CPU only, execute the code for the CPU (DualSPHysics4.2CPU.exe). This means the GPU is not required, which means CUDA is not needed.
Sign In or Register to comment.