• Study increasingly sophisticated parallel merge kernels • Observe the combined effects of data - dependent execution and a lack of data parallelism on GPU algorithm design

电子科技大学：《GPU并行编程 GPU Parallel Programming》课程教学资源（课件讲稿）Lecture 07 JOINT CUDA-MPI PROGRAMMING

parallel histogram Data Racing condition privatized histogram kernel

Memory access efficiency Tiled Matrix Multiplication Tiled Matrix Multiplication Kernel Handling Boundary Conditions in Tiling Tiled Kernel for Arbitrary Matrix Dimensions

Introduction to Heterogeneous Parallel Computing CUDA C vs. CUDA Libs vs. OpenACC Memory Allocation and Data Movement API Functions Data Parallelism and Threads

《GPU并行编程 GPU Parallel Programming》课程教学资源（参考文献）NVIDIA CUDA C Programming Guide（Design Guide，June 2017）

《GPU并行编程 GPU Parallel Programming》课程教学资源（参考文献）NVIDIA Parallel Prefix Sum（Scan）with CUDA（April 2007）

《GPU并行编程 GPU Parallel Programming》课程教学资源（参考文献）Program Optimization Space Pruning for a Multithreaded GPU

《GPU并行编程 GPU Parallel Programming》课程教学资源（参考文献）Some Computer Organizations and Their Effectiveness

《GPU并行编程 GPU Parallel Programming》课程教学资源（参考文献）An Asymmetric Distributed Shared Memory Model for Heterogeneous Parallel Systems