Parallel and High Performance Programming with Python, 2/e

Book Description

Python is the backbone for data science, AI, and cloud computing, and the demand for speed and scalability has never been higher. That’s why mastering parallel and high-performance programming is essential to transform Python into a tool that meets modern performance demands.

Building on the success of the first edition, Parallel and High Performance Programming with Python (2nd Edition) expands and modernizes the original work, adding new frameworks, deployment patterns, and acceleration techniques for next-generation computing.

You’ll begin by mastering the core concepts of parallelism, threading, and multiprocessing, then move into asynchronous programming for responsive and efficient workloads. The book guides you through distributed Python across clusters, followed by deep dives into GPU acceleration using CUDA and PyTorch. You’ll explore real-world applications in data science and artificial intelligence, and learn how to scale pipelines seamlessly with Ray, Modin, and Spark.

Advanced chapters introduce Joblib optimization, Kubernetes, and serverless scaling for cloud-native workloads, and cutting-edge topics such as FPGA acceleration and quantum computing, giving you a future-ready performance toolkit. Packed with hands-on examples, benchmarks, and deployment-ready best practices, this second edition helps you turn everyday Python into a high-performance, production-grade system.

What you will learn
● Design and optimize high-performance Python applications using parallelism, concurrency, and async patterns.
● Profile, diagnose, and eliminate CPU, I/O, and memory bottlenecks for real-world workloads.
● Accelerate compute-intensive tasks using CUDA kernels, PyTorch tensors, NumPy vectorization, and GPU-enabled deep learning workflows.
● Build and scale distributed systems seamlessly with Dask, Ray, Apache Spark, and Modin for massive data processing.
● Deploy and orchestrate compute pipelines on Kubernetes, AWS Lambda, and Azure Functions for cost-efficient scalability.
● Integrate advanced acceleration technologies like Joblib, FPGA workflows, and quantum simulation frameworks to stay ahead of the curve.

Who is this book for?
This book is tailored for data scientists, machine learning engineers, Python developers, and backend programmers who need to boost performance and scalability in their applications. It also serves data and cloud engineers, DevOps/MLOps professionals, HPC specialists, and distributed systems developers building high-throughput pipelines. Technical architects, research engineers, and quantitative analysts will find advanced design patterns to optimize workloads across CPUs, GPUs, clusters, FPGAs, and serverless or quantum environments.

Table of Contents
1. Introduction to Parallel Programming
2. Parallel Programming with Threads
3. Parallel Programming with Processes
4. Asynchronous Programming
5. Distributed Python
6. GPU Programming with Python
7. Parallel Computing Applications
8. Parallel Computing for Data Science
9. Parallel Computing for Artificial Intelligence
10. Future of Parallel Programming
11. Modern Frameworks for Parallelism and Distribution: Ray and Modin
12. PySpark
13. Joblib
14. Parallelization on Cloud and Serverless Systems
15. Parallel Programming with FPGAs
16. Introduction to Quantum Co

About the Author

Fabio Nelli holds a Master's Degree in Chemistry and a Bachelor's Degree in IT and Automation Engineering. He currently works at various research institutes and private companies, where he delivers educational courses on data analysis and data visualization technologies. He contributes to writing articles on the web and writes in-depth books on the subject.