Gpu Parallel Program Development Using Cuda

ISBN-13: 9780367572242 / Angielski / Miękka / 2020 / 440 str.

Tolga Soyata

Opis

This book teaches GPU programming by introducing CPU multi-threaded programming and bases GPU massively-parallel programming on this foundation. The differences among families of GPUs are also studied. The book also explores CUDA libraries, OpenCL, GPU programming with other languages and API libraries, and the deep learning library cuDNN

Spis treści

Introduction to Parallelism and GPU Programming. Introduction to Pthreads. Introduction to Multi-threading. Multi-threading Analysis. Introduction to AWS and the first GPU program. Analysis of the first CUDA program. How to run programs outside Amazon, on a different Unix system. How to run programs outside Amazon, on a different Windows system. Introduction to GPU Multi-threading and Global Memory. In-depth look at Global Memory and GPU Threads. Introduction to Shared Memory. In-depth look at Shared Memory. Performance Bottlenecks in GPU Architecture. Compute Capability and PTX. Host-Device Data Movement Mechanism. CUDA Streaming. CUDA Atomics. OpenGL-CUDA Interoperability. OpenGL and Real-Time GPU Processing. Example Applications - Edge Detection (CUBLAS). Example Applications - Sophisticated Image Processing (CUFFT). Example Applications- Elastography Based Tumor Detection (CUSOLVER). Example Applications - Real-Time Face Recognition (OpenCV, CUBLAS).

O autorze

Tolga Soyata is an associate professor in the Electrical and Computer Engineering department of SUNY Albany.