Overview
Explore the concept of "cache friendliness" in data structures and algorithms through this comprehensive conference talk from code::dive 2019. Delve into the importance of cache memory in modern CPUs and its impact on performance. Learn about the properties of cache memory, including its speed advantages and size limitations. Discover how to analyze cache efficiency, improve data structure layout in memory, and optimize access patterns for better performance. Gain insights into writing fast software by default and techniques for further improvements. Follow along as the speaker, Björn Fahller, an experienced software architect and programmer, demonstrates tools for analyzing cache efficiency and provides practical examples of problematic constructions and their better alternatives. Develop an intuition for creating high-performance software while covering topics such as linear arrays, binary search, map performance, cancellation, heaps, and B-heaps.
Syllabus
Intro
Background
Cache model
Valgrind
Cache Simulator
Cache miss analysis
Timer structure
Perfstats
Linear Array
Binary Search
Map
Map performance
Cancellation
Heaps
B Heap
Taught by
code::dive conference