How best to implement the interaction of Cuda and OpenGL? I need to make calculations on Cuda and somehow draw them.
I am not a professional, I can make a mistake in technology, can you tell me how to do this better?
1 answer
Yes it is possible. You need to learn CUDA / OpenGL Interoperability:
- https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__OPENGL.html
- http://on-demand.gputechconf.com/gtc/2012/presentations/S0267A-Mixing-Graphics-and-Compute-with-Multiple-GPUs-Part-A.pdf .
If you just need to perform calculations (CUDA is not critical) and draw, then you can still explore one of the following approaches:
- OpenCL / OpenGL Interoperability: https://software.intel.com/en-us/articles/opencl-and-opengl-interoperability-tutorial
- Compute Shader: https://www.khronos.org/opengl/wiki/Compute_Shader
- Vulkan: http://www.duskborn.com/a-simple-vulkan-compute-example/
Unick unick
1.556 5 silver marks 12 bronze marks
|