In this chapter, we looked at how to integrate GPGPU processing into a C++ application in the form of OpenCL. We also looked at the GPU memory hierarchy and how this impacts performance, especially in terms of host-device communication.
You should now be familiar with GPGPU implementations and concepts, along with how to create an OpenCL application, and how to compile and run it. How to avoid common mistakes should also be known.
As this is the final chapter of this book, it is hoped that all major questions have been answered, and that the preceding chapters, along with this one, have been informative and helpful in some fashion.
Moving on from this book, the reader may be interested in pursuing any of the topics covered in more detail, for which many resources are available both online and offline. The topic of multithreading and related areas is very large and touches...
