Efficient ML Computing

Abstract

Efficient ML Computing offers readers an entry point to understand comprehensive machine learning systems. As resource-constrained edge computing sees rapid expansion, the ability to construct efficient ML pipelines grows crucial. This book aims to demystify the process of developing complete ML systems suitable for deployment - spanning key phases like data collection, model design, optimization, acceleration, security hardening, and integration. The text touches on the full breadth of concepts relevant to general ML engineering across industries and applications. Readers will learn basic principles around designing ML model architectures, hardware-aware training strategies, performant inference optimization, benchmarking methodologies and more. Additionally, crucial systems considerations in areas like reliability, privacy, responsible AI, and solution validation are also explored in depth. In summary, the book strives to equip newcomers and professionals alike with integrated knowledge covering full stack ML system development and universal concepts required to unlock production ML.

Preface

Welcome to Efficient ML Computing. This book is your gateway to the fast-paced world of AI systems. It is a fork of the open source book from CS249r at Harvard University.

The aim of the original open-source book is a collaborative effort that brings together insights from students, professionals, and the broader community of applied machine learning practitioners. It is intended to be a one-stop guide that dives deep into the nuts and bolts of AI systems and their many uses.

The customized book is the companion texbook for UCI’s Efficient ML Computing class, and the aim is to provide students with the foundational knowledge to start research in this exciting and fast paced area.

The book covers principles, algorithms, and real-world application case studies, aiming to give you a deep understanding that will help you navigate the ever-changing landscape of AI. By keeping it open, we’re not just making learning accessible; we’re inviting new ideas and ongoing improvements. In short, we’re building a community where knowledge is free to grow and light the way forward in global embedded AI tech.

What You’ll Need to Know

You don’t need to be a machine learning whiz to dive into this book. All you really need is a basic understanding of systems and a curiosity to explore how hardware, AI, and software come together. This is where innovation happens, and a basic grasp of how systems work will be your compass.

We’re also focusing on the exciting overlaps between these fields, aiming to create a learning environment where traditional boundaries fade away, making room for a more holistic, integrated view of modern tech. Your interest in AI and low-level software will guide you through a rich and rewarding learning experience.

Book Conventions

For details on the conventions used in this book, check out the Conventions section.

Get in Touch

Got questions or feedback? For the customized content, feel free to [e-mail Prof. Thomas Yeh] directly, or you are welcome to start a discussion thread on GitHub.

Contributors

A big thanks to everyone who’s helped make this book what it is! You can see the full list of individual contributors here and additional GitHub style details here.