Address-Space Qualification | Address space restriction specifiers for pointers |
Assertions | Enhanced CUDA assertions, including: |
▼Device Functions | Device-side APIs |
Atomics | Type-safe wrappers around the atomic operators found in GPU runtime APIs |
▼Device Intrinsics | These APIs provide direct (and usually non-portable) access to hardware features that are either not supported by the compiler (yet), or which manufacturers don't document |
Texture Intrinsics | Plugging some holes in the NVIDIA® texture fetch API, and making it less annoying to use in template code |
Lanemask Functions | Functions for computing lane masks |
Shuffle | Generic warp shuffle support supporting arbitrary input types |
Synchronisation Functions | Access to the more elaborate syncthreads variants |
▼Host API | Host-side CUDA API (use instead of libcuda) |
Exceptions | Exception objects for errors from the GPU runtime API |
NVTX Tools Extension | A (very cut-down) API wrapper for NVTX |
Named Requirements | Similarly to the standard library, we define Named Requirements to specify constraints that must be satisfied by type parameters |
▼Utilities | |
▼Memory Manipulation Utilities | |
▼Smart Pointer Utilities | Utilities for dealing with smart pointers |
Deleters | Smart pointer deleter functions, suitable for use with STL smart pointers |
Type Traits |