• Advertisement

Mona2000

Member
  • Content count

    300
  • Joined

  • Last visited

Community Reputation

1967 Excellent

About Mona2000

  • Rank
    Member
  1. Obviously the highly specific technique for that use case is gonna be faster than the generic instancing one that doesn't allow further optimization and you should use it whenever possible, but that's not a valid comparison to show instancing overhead.
  2. It performs identically on the GPU and orders of magnitude faster on the CPU. How is that not better?
  3. DX12 High precision texture.

      Well, DDS is going to save you loading time..
  4.   I think the documentation is wrong, Jesse Natalie (the user who replied) is working on DirectX at Microsoft..
  5. What to do besides fencing?

      That is meaningless, 15 seconds on what? A quad core i7? A Pentium 2? A 16 core Xeon?
  6. I don't see any effects extensions being used in your code and you don't seem to be compiling with an effects target (fx_*). Can we see the call to D3DCompileFromFile?
  7. getting a new PC for DX development

    That's not how you're supposed to use an SSD, though. You have to put everything on it for it to make a real difference: operating system, applications, user data, games (modern games are 40+GB btw), development assets... I was a SSD skeptic like you until the moment I tried one.
  8. Isn't that kind of optimization performed on the IL? Both clip and discard emit a discard_nz.
  9. getting a new PC for DX development

    Also your development mouse shouldd not have a scroll wheel because some of your players might not have it as well.
  10. Matrix 16 byte alignment

      The GNU libc guarantees 16-byte alignment on x64 (http://www.gnu.org/software/libc/manual/html_node/Aligned-Memory-Blocks.html).
  11. I just tried it and the compiler (fxc.exe) emits a mad for a * b + c if you don't force IEEE strictness (/Gis) or disable optimizations (/Od).
  12. Both OpenGL and DirectX are left-handed in clip space. You're probably thinking of world space and view space in the fixed function pipeline, which were right handed in OpenGL. Nowadays the FFP is dead and handedness in those spaces is up to the user.
  13. Can you even connect a monitor to the NVIDIA card in an Optimus setup? I thought the outputs were always connected to the IGP and the NVIDIA card works by writing into the IGP framebuffer.
  14. The only thing that changes is how you bind it (with *SSetConstantBuffers1 instead of *SSetConstantBuffers). See https://msdn.microsoft.com/en-us/library/windows/desktop/hh404649(v=vs.85).aspx
  15. It fails because the NVIDIA card doesn't have any outputs. It's how those dual GPU setups work, the final output has to go through the integrated card no matter what. Both NVIDIA and AMD provide ways to force using the dedicated card; the easiest one is adding this to your program: extern "C" { __declspec(dllexport) DWORD NvOptimusEnablement = 1; __declspec(dllexport) int AmdPowerXpressRequestHighPerformance = 1; }
  • Advertisement