Overview
Explore a comprehensive conference talk on implementing high-performance hash tables in C++. Delve into the success story of creating world-class hash tables using Robin Hood hashing techniques and parallel computation. Learn how to maintain the Robin Hood hash table invariant while achieving optimal performance through normal integer operations. Discover the benefits of this approach, including improved cache locality and reduced false matches. Examine real source code, benchmarks, and Compiler Explorer disassembly to understand the composable parallel programming abstraction layer. Gain insights into how this implementation outperforms alternative schemes and provides a more understandable codebase. Presented by Eduardo Madrid and Scott Bruce at CppCon 2022, this talk offers valuable knowledge for C++ developers seeking to enhance their understanding of efficient data structures and parallel programming techniques.
Syllabus
Implementing Understandable World Class Hash Tables in C++ - Eduardo Madrid, Scott Bruce CppCon 2022
Taught by
CppCon