Chapter 7. Heterogeneous and Distributed Computing

A computation expressed using TensorFlow can be executed with little or no changes on a wide variety of heterogeneous systems, ranging from mobile devices such as phones and tablets up to large-scale distributed systems of hundreds of machines and thousands of computational devices, such as GPU cards.

In this chapter, we explore this fundamental topic on TensorFlow. In particular, we shall consider the possibility of executing TensorFlow models on GPU cards and distributed systems.

GPUs have additional advantages over CPUs, including having more computational units and having a higher bandwidth for memory retrieval. Furthermore, in many deep learning applications that require a lot of computational effort, GPU graphics specific capabilities can be exploited to further speed up calculations.

At the same time, a distributed computing strategy can be useful if you have to handle a very large dataset to train your model.

The chapter introduces...