CUDA C++编程手册（总论）

CUDA C++编程手册（总论）
CUDA C++ Programming Guide
The programming guide to the CUDA model and interface.
Changes from Version 10.0
Use CUDA C++ instead of CUDA C to clarify that CUDA
C++ is a C++ language extension not a C language.
General wording improvements throughput the guide.
Fixed minor typos in code examples.
Updated From Graphics Processing to General Purpose Parallel Computing.
Added reference to NVRTC in Just-in-Time Compilation.
Clarified linear memory address space size in Device Memory.
Clarified usage of CUDA_API_PER_THREAD_DEFAULT_STREAM
when compiling with nvccin Default Stream.
Updated Host Functions (Callbacks) to use cudaLaunchHostFunc instead of the deprecated cudaStreamAddCallback.
Clarified that 8 GPU peer limit only applies to
non-NVSwitch enabled systems in Peer-to-Peer Memory Access.
Added section IOMMU on Linux.
Added reference to CUDA Compatibility in Versioning and
Compatibility.
Extended list of types supported by __ldg() function in Read-Only Data Cache Load Function.
Documented support for unsigned short int with atomicCAS().
Added section Address Space Predicate Functions.
Added removal notice for deprecated warp vote
functions on devices with compute capability 7.x or higher in Warp Vote Functions.
Added documentation for __nv_aligned_device_malloc() in Dynamic Global Memory Allocation and Operations.
Added documentation of cudaLimitStackSize in CUDA Dynamic Parallelism Configuration Options.
Added synchronization performance guideline to CUDA
Dynamic Parallelism Synchronization.
Documented