GTC 2009 OpenGL Gold
Download as PPT, PDF2 likes894 views
Michael I Gold discusses interoperability between various graphics APIs such as CUDA, OpenGL, OpenCL, and Direct3D. This allows sharing resources such as buffers and textures between APIs and contexts. CUDA and OpenGL can share buffer objects where CUDA kernels can directly access OpenGL buffer memory. OpenCL also allows creating OpenCL memory objects from textures and buffers in OpenGL and other contexts. Direct3D resources can be imported into OpenGL on Windows XP. Overall, interoperability provides mechanisms for sharing data across APIs and heterogeneous systems.
GTC 2009 OpenGL Gold
- 1. Michael I Gold NVIDIA Corporation OpenGL Interoperability
- 2. What is Interoperability? Share memory objects between APIs or contexts within an API Share within a device, or between devices Implicit or explicit synchronization
- 3. Who can play? CUDA OpenCL OpenGL Direct3D
- 4. CUDA Interoperability Import OpenGL buffer objects as CUDA device pointers Read or write buffer memory directly from a kernel Implicit synchronization within a CPU thread Texture support coming soon
- 5. CUDA Interop: Example Application OpenGL driver CUDA driver Quadro or GeForce display Tesla or GeForce Interop fast Multi-card interop fastest with GL on Quadro. Transfer between cards w/o CPU intervention GeForce performs indirect copy, functional but slower Interop faster
- 6. CUDA example: Initalization // Create a GL buffer object glBindBuffer(GL_PIXEL_UNPACK_BUFFER, buffer); glBufferData(GL_PIXEL_UNPACK_BUFFER, buffer_size, NULL, GL_STREAM_COPY); // Register the buffer object for CUDA interop cuGLRegisterBufferObject(buffer); cuGLSetBufferObjectMapFlags(buffer, CU_GL_MAP_RESOURCE_FLAGS_WRITE_DISCARD);
- 7. CUDA example: Per-frame usage // map the buffer, invoke a kernel to populate it, and release it back to OpenGL cuGLMapBufferObjectAsync(&dptr, &size, buffer, stream); call_kernel(dptr, size); cuGLUnmapBufferObjectAsync(&dptr, &size, buffer, stream); // Use the buffer as a pixel buffer object to populate a texture glTexSubImage2D( … , NULL);
- 8. CUDA resources Drivers, Toolkit and SDK http:// www.nvidia.com/cuda Forums http:// forums.nvidia.com
- 9. OpenCL Interoperability Analogous to CUDA Create cl_mem from textures and buffers Requires explicit synchronization
- 10. OpenCL Shared Objects OpenCL 3D image object OpenGL renderbuffer OpenGL buffer OpenCL buffer object OpenGL texture 2D (or cubemap face) OpenGL texture 3D OpenCL 2D image object OpenCL 2D image object clCreateFromGLBuffer clCreateFromGLTexture2D clCreateFromGLTexture3D clCreateFromGLRenderbuffer OpenGL OpenCL
- 11. OpenGL Interoperability Copy between image objects 3D memcpy No format conversions Copy between dimensionalities E.g. A slice of a 3D texture and a face of a cubemap are treated equally Copy between GPUs Complements WGL_NV_gpu_affinity
- 12. OpenGL Interop: Example Application OpenGL driver OpenGL driver Quadro display Quadro Interop fast Quadro only Transfer between cards, no host copy Rendering window constrained to affinity GPU GPU affinity Offscreen buffer
- 13. OpenGL Interop: Usage wglCopyImageSubDataNV( hSrcRC, srcName, srcTarget, srcLevel, srcX, srcY, srcZ, hDstRC, dstName, dstTarget, dstLevel, dstX, dstY, dstZ, width, height, depth);
- 14. OpenGL Interop: Resources NV_copy_image extension specification http:// developer.download.nvidia.com / opengl/specs/GL_NV_copy_image.txt NV_gpu_affinity extension specification http:// developer.download.nvidia.com / opengl/specs/WGL_NV_gpu_affinity.txt
- 15. Direct3D Interoperability Import Direct3D buffer and image resources into OpenGL Vertex and Index buffers -> buffer objects Textures and surfaces -> textures and renderbuffers Currently WinXP / Direct3D 9 only Experimental extension WGL_ NVX _DX_interop
- 16. Direct3D Interop: Example Application OpenGL driver DX driver Quadro or GeForce display Interop fast OGL can map DX buffers, not the other way Single GPU only (at this time)