CUDA 13: High-Performance Computation and Graphics

Unlock high‑performance computing with CUDA 13Nvidia's GPUs offer a stunning amount of computational power, but harnessing that power requires the right tools. Whether you're crunching numbers, filtering images, or rendering graphics, CUDA will help you dramatically improve performance. Any programmer familiar with C or C++ can use CUDA to tackle today’s hardest computational problems.CUDA 13 explores the latest version of CUDA, and shows how to write applications that leverage this incredible toolset. This book walks through the entire development process, from installing the toolkit to compiling, linking, and debugging applications. The twenty-nine example projects cover a vast range of topics, including matrix operations, frequency analysis, image filtering, and three-dimensional rendering.Matthew Scarpino has been an author, programmer, and engineer for over twenty years. GPU programming is one of his favorite topics, and he's the author of The WebGPU Sourcebook and OpenCL in Action.Table of ContentsChapter 1: IntroductionChapter 2: The CUDA ToolkitChapter 3: Programming FundamentalsChapter 4: Kernel DevelopmentChapter 5: Advanced Development TopicsChapter 6: Accelerating Linear Algebra with cuBLASChapter 7: Solving Systems of Equations with cuSOLVER and cuDSSChapter 8: Frequency Analysis with cuFFTChapter 9: Processing Images with NPPIChapter 10: Graphical Rendering with OpenGLChapter 11: CUDA-OpenGL InteroperabilityChapter 12: Processing Textures and SurfacesAppendix A: The CUDA Driver API and Runtime CompilationAppendix B: Parallel Thread Execution (PTX)All example code is freely available for download:Windows: https://github.com/mattscar/cuda-book-winLinux: https://github.com/mattscar/cuda-book-gnu Read more