Programming massively parallel processors : a hands-on approach / Print
David B. Kirk and Wen-mei W. Hwu.
- Second edition.
- xx, 496 pages : illustrations ; 23 cm
Previous ed.: 2010.
Includes bibliographical references and index.
History of GPU computing -- Introduction to data parallelism and CUDA C -- Data-parallel execution model -- CUDA memories -- Performance considerations -- Floating-point considerations -- Parallel patterns : convolution -- Parallel patterns : prefix sum -- Parallel patterns : sparse matric-vector multiplication -- Application case study : advanced MRI reconstruction -- Application case study : molecular visualization and analysis -- Parallel programming and computational thinking -- An introduction to OpenCL -- Parallel programming with OpenACC -- Thrust : a productivity-oriented library for CUDA -- CUDA FORTRAN -- An introduction to C++ AMP -- Programming a heterogeneous computing cluster -- CUDA dynamic parallelism -- Conclusion and future outlook.