Explore crucial aspects of multithreaded performance optimization in this conference talk from Meeting C++ 2023. Dive into false sharing and hardware destructive interference, understanding their impact on cache efficiency and application performance. Learn to identify and mitigate false sharing issues that degrade multithreaded performance. Gain insights into aligning data structures based on hardware destructive interference for optimal efficiency. Through real-world examples and performance benchmarks, discover practical strategies to overcome cache-related bottlenecks and maximize the performance of multithreaded applications. Essential for software developers, system architects, and researchers looking to unlock the full potential of their concurrent systems.
Optimizing Multithreading Performance - False Sharing and Hardware Destructive Interference
Meeting Cpp via YouTube
Overview
Syllabus
Optimizing Multithreading Performance - Shivam Kunwar - Meeting C++ 2023
Taught by
Meeting Cpp