Table of Contents
- 1. Difference between the driver and runtime APIs
- 2. API synchronization behavior
- 3. Stream synchronization behavior
- 4. Modules
- 4.1. Data types used by CUDA driver
- 4.2. Error Handling
- 4.3. Initialization
- 4.4. Version Management
- 4.5. Device Management
- 4.6. Device Management [DEPRECATED]
- 4.7. Primary Context Management
- 4.8. Context Management
- 4.9. Context Management [DEPRECATED]
- 4.10. Module Management
- 4.11. Memory Management
- 4.12. Unified Addressing
- 4.13. Stream Management
- 4.14. Event Management
- 4.15. Execution Control
- 4.16. Execution Control [DEPRECATED]
- 4.17. Occupancy
- 4.18. Texture Reference Management
- 4.19. Texture Reference Management [DEPRECATED]
- 4.20. Surface Reference Management
- 4.21. Texture Object Management
- 4.22. Surface Object Management
- 4.23. Peer Context Memory Access
- 4.24. Graphics Interoperability
- 4.25. Profiler Control
- 4.26. OpenGL Interoperability
- 4.27. Direct3D 9 Interoperability
- 4.28. Direct3D 10 Interoperability
- 4.29. Direct3D 11 Interoperability
- 4.30. VDPAU Interoperability
- 4.31. EGL Interoperability
- 5. Data Structures
- 5.1. CUDA_ARRAY3D_DESCRIPTOR
- 5.2. CUDA_ARRAY_DESCRIPTOR
- 5.3. CUDA_LAUNCH_PARAMS
- 5.4. CUDA_MEMCPY2D
- 5.5. CUDA_MEMCPY3D
- 5.6. CUDA_MEMCPY3D_PEER
- 5.7. CUDA_POINTER_ATTRIBUTE_P2P_TOKENS
- 5.8. CUDA_RESOURCE_DESC
- 5.9. CUDA_RESOURCE_VIEW_DESC
- 5.10. CUDA_TEXTURE_DESC
- 5.11. CUdevprop
- 5.12.
- 5.13. CUipcEventHandle
- 5.14. CUipcMemHandle
- 5.15. CUstreamBatchMemOpParams
- 6. Data Fields
- 7. Deprecated List
- Notices