Machine Learning in Production: Developing and Optimizing Data Science Workflows and Applications

15. Hardware Fundamentals

15.1 Introduction

To get the most out of your machines, it is important to understand the capabilities and limitations of them both physically and abstractly. In this chapter, we will go over several core hardware concepts. With an adequate understanding of these concepts, you can use your resources as effectively as possible, avoiding common (and uncommon) pitfalls.

15.2 Random Access Memory

Random access memory exists on the vast majority of, if not all, modern computers. It’s fast to access, so it’s commonly used as scratch paper when processes are performing tasks. In algorithm work, this is commonly referred to as working memory, or working tape.

From the perspective of the data scientist or engineer, there ...