Book Image

Mastering C++ Multithreading

By : Maya Posch
Book Image

Mastering C++ Multithreading

By: Maya Posch

Overview of this book

Multithreaded applications execute multiple threads in a single processor environment, allowing developers achieve concurrency. This book will teach you the finer points of multithreading and concurrency concepts and how to apply them efficiently in C++. Divided into three modules, we start with a brief introduction to the fundamentals of multithreading and concurrency concepts. We then take an in-depth look at how these concepts work at the hardware-level as well as how both operating systems and frameworks use these low-level functions. In the next module, you will learn about the native multithreading and concurrency support available in C++ since the 2011 revision, synchronization and communication between threads, debugging concurrent C++ applications, and the best programming practices in C++. In the final module, you will learn about atomic operations before moving on to apply concurrency to distributed and GPGPU-based processing. The comprehensive coverage of essential multithreading concepts means you will be able to efficiently apply multithreading concepts while coding in C++.
Table of Contents (17 chapters)
Title Page
Credits
About the Author
About the Reviewer
www.PacktPub.com
Customer Feedback
Preface
8
Atomic Operations - Working with the Hardware

GPGPU and multithreading


Combining multithreaded code with GPGPU can be much easier than trying to manage a parallel application running on an MPI cluster. This is mostly due to the following workflow:

  1. Prepare data: Readying the data which we want to process, such as a large set of images, or a single large image, by sending it to the GPU's memory.
  2. Prepare kernel: Loading the OpenCL kernel file and compiling it into an OpenCL kernel.
  3. Execute kernel: Send the kernel to the GPU and instruct it to start processing data.
  4. Read data: Once we know the processing has finished, or a specific intermediate state has been reached, we will read a buffer we passed along as an argument with the OpenCL kernel in order to obtain our result(s).

As this is an asynchronous process, one can treat this as a fire-and-forget operation, merely having a single thread dedicated to monitoring the process of the active kernels.

The biggest challenge in terms of multithreading and GPGPU applications lies not with the host...