A modern CPU is a combination of many sophisticated features. Among them, the memory hierarchy probably has the most impact on the performance of your programs. Understanding how it works is essential for writing high-performance applications.
The talk provides an overview of the structure and algorithms of the memory hierarchy in a modern CPU. Specific topics include cache organization, locality principle and cache-friendly data organization. In addition, some practical guidelines are given to avoid common performance traps.
The talk does not require any prior knowledge of hardware design.