Exercises

CUDA programming I

Compiling: compiling
First simple CUDA program: first-cuda-program

Tools

Profiling with nvvp: profile-nvvp
Profiling with nvprof: profile-nvprof
Debugging with cuda-memcheck: debug-cuda-memcheck
Debugging with nsight: debug-nsight

CUDA programming II

Error checking: error-checking
Jacobi iteration: jacobi

CUDA programming III

Timing using events: jacobi-timing-events
Using streams: simple-streams

Kernel optimization

Kernel optimization: kernel-optimization

CUDA programming IV

First unified memory program: first-unified-memory-program
Using unified memory and streams: unified-memory-streams

Dynamic Parallelism

Synchronization: dynamic-sync
Mandelbrot set: dynamic-mandelbrot

Multi-GPU programming

Multi-GPU programming: multi-GPU

MPI and CUDA

MPI Ping Pong: mpi-ping-add-pong