Hands-On GPU Computing with Python Paperback – 13 May 2019

The circumstances under which this book was written weren't very convenient because it was not supposed to be published in May 2019 but actually a month later. Advancing it by a month made it very difficult to cover each and every aspect in detail(as planned earlier). Even then, I've tried my best to keep up with the expected quality.It is sad that the know-how of assembling PC hardware is still not seen as an essential part of GPU Computing. Programming and Computing are NOT synonymous with each other, which is why the book is not titled "Hands-On GPU Programming with Python". For this reason, the differences between Programming and Computing have been specifically highlighted and discussed when chapter 4 begins.The rating I have given is an unbiased one, not because I wrote it within just 2 months, but because it is an honest and original effort from a PhD candidate trying to understand biological problems through applied computing. Hence, the last chapter focuses on computational biochemistry as a real-world example. I've tried to educate readers at all proficiency levels. Therefore, there is something for everyone to gain from this book.If you are interested in learning the core essentials of GPU Programming, then please do not buy this book.If you want to know what a GPU is, how to build a PC with one and use it to solve real-world computational problems through specifically programmable GPU enabled code, then this book is for you.

Most of the developers, data-scientists, researchers, etc. don't realize the importance of gaining knowledge about GPUs and their computational efficiencies. The author has explained everything required from ground-zero to the top level, in an optimized approach to suit the interests varying from that of a naive reader to that of a novice. I loved how the author has emphasized on educating the reader about the basics of building an optimal system fro GPU computing, in adherent to various types of budget.Most importantly, the book consists of all the minute information required to get comfortable with Nvidia's CUDA and AMD's ROCm platform for GPU Computing. There are detailed examples of how to install and implement various types of packages and in-built libraries to get started with harnessing the power of parallel computing with GPUs. Especially relevant, there are assignments at the end of each practical chapters, which the readers can salvage to become experts in using GPUs for computational needs. The author has also introduced certain sections in the book to indulge the reader about the power and advent of GPUs in research, hence only to excite and get them the feel of it.The elucidation of GPU Computing with TensorFlow, PyTorch, etc. in the book is something the reader will automatically grind for. Nonetheless, the use of DeepChem (Deep Learning in Chemistry), and it's basic implementation with GPU Computing, was the ultimate chapter for me. Overall, I had an amazing experience, and shall still refer it in the future, for my upcoming projects.

This is a book about GPU computing, NOT programming.No GPU/Python programming concepts and skills are shown.Most importantly, this book lacks in focus. Too much facts,but no much useful information, concepts or skills are introduced.The first 30% of the book is allocated to the history of CPU and GPU.Starting from 4004 CPU from early days... Although the history is clear and well written,it does not help in improving your GPU computing or programming work.Readers better skip the first 30%.Furthermore, there is no need to list different combinations of building a computer withIntel CPU, AMD CPU, Nvdia GPU, Radeon, at entry-level, at middle level and at high-end level.CUDA, PyCUDA and other modules are included but explained briefly with 2 to 4 paragraphs.Some paragraphs are dedicated to help the reader install the relevant libraries.Some reader may find it helpful. There are examples at the end of the book demonstrating Pytorchand another module. They are practical.If you are are looking for introduction to GPU computing,there are other short and good books which help you to buildup concepts about GPU computing and machine learning.If you are looking for concepts and skills in improving your programming.It is NOT the book for you.Better have a look at a sample of this book and realize what I am talking.