CUDA NVRTC & CDP samples

Feature list:

• CUDA Driver API
• CUDA Runtime Compilation
• CUDA Dynamic Parallelism
• saxpy, qsort, mandelbrot samples
• pageable vs pinned memory sample - transfer times captured with CUDA events

Requirements

• CMake
• NVRTC requires CUDA GPU sm20+
• CDP requires CUDA GPU sm35+
• Windows only

Mandelbrot set with CUDA Dynamic Parallelism - The Mariani-Silver Algorithm

References:

• NVIDIA CUDA SDK samples
• NVRTC saxpy sample - CUDA Toolkit Documentation
• NVRTC cdp sample - CUDA Toolkit Documentation
• Adaptive Parallel Computation with CDP - Parallel Forall
• CUDA Dynamic Parallelism API and Principles - Parallel Forall
• How to Optimize Data Transfers in CUDA C/C++ - Parallel Forall