Parallel MATLAB for Multicore and Multinode Computers

Parallel MATLAB: A Practical Guide to Multicore and Multinode Computing

MATLAB is a powerful environment for numerical computing, but as problems grow, a single processor core quickly becomes a bottleneck. Parallel MATLAB for Multicore and Multinode Computers shows scientists and engineers how to adapt MATLAB code to run efficiently on modern parallel hardware—from laptops with multiple cores to clusters with hundreds of nodes.

Written by Jeremy Kepner of MIT Lincoln Laboratory, this book introduces pMatlab, a library that extends MATLAB with parallel data structures and functions. Through clear explanations and practical examples, you’ll learn to write parallel programs that scale.

What the Book Covers

The book is organized into three parts:

• Fundamentals: A primer on notation and interfaces, an introduction to pMatlab, and techniques for interacting with distributed arrays.
• Advanced Techniques: Parallel programming models (manager/worker, message passing, distributed arrays), advanced distributed array programming, and performance metrics and software architecture.
• Case Studies: Detailed analyses of parallel applications—Stream, RandomAccess, Fast Fourier Transform, and High Performance Linpack—demonstrating how to apply the concepts to real-world problems.

An appendix covers notation for hierarchical parallel multicore algorithms, making the book a complete reference for parallel MATLAB programming.

Who This Book Is For

Parallel MATLAB for Multicore and Multinode Computers is written for scientists, engineers, and graduate students who are comfortable with MATLAB and want to leverage parallel computing. Whether you’re working on data-intensive simulations, signal processing, or numerical algorithms, this book provides the tools and insights to accelerate your work.

About the Author

Jeremy Kepner is a researcher at MIT Lincoln Laboratory, where he works on high-performance computing and parallel programming. His expertise shines through in the book’s practical approach and deep technical understanding.

Why It Matters

As multicore processors and distributed systems become the norm, parallel programming skills are increasingly essential. This book bridges the gap between MATLAB’s ease of use and the performance demands of modern computational science. By the end, you’ll be equipped to write parallel MATLAB code that runs faster and scales further.

Add this ebook to your library and start writing faster, more scalable MATLAB code today.