About

Who should read

This book is addressed to all developers who would like to go into the world of CUDA–GPU processing but ones not sure about they capability of“low level programming”–memory allocations, pointers etc.

There are 2 parts in this book:

• Part 1: where you can learn how from language like java access CUDA libraries and start doing real parallel computing on hundreds of processors

• Part 2: where you’ll learn how to not only use GPU but how to get full advantage of graphics hardware: use internal memory/ registers, cache, streaming.

• Part 3: as Guy Kawasaki said in one of his lectures:”you should under-promise but always over-deliver”so this is bonus part for your consideration: where could you go next, what currently(2015) CUDA world is passionate about

Side note: I’m writing this in May 2015–and during the course of this book you’ll see occasional recurrence about time, that because you will be suggested to use libraries and those libraries are changing with time. If you’re developer you probably know it already - at the moment I’m working on scala and every single book that you can buy was at least written 6 months ago…so if you want to have working examples in scala–you need to find what was change during that period.

When I started this book CUDA version was 6.5, currently it is 7.0.

Scope

Purpose of this book is:

Easily switch you into CUDA world from high level language like Java

Show you what all this GPU stuff is about

Introduce you to existing/ popular libraries (2015)

Hopefully seed you with GPU craziness–low level hardcore hardware programming

Show you wide range of possibilities that GPU programming could give us

We are living in world of technology, where every year we are seeing enormous changes that ten years ago look like song of SF–I believe GPU programing could take big part in that revolution.