Explore caching in modern CPUs and its relationship with Rust data types and structures in this 53-minute talk from Rust Cologne. Gain insights into efficient code writing that maximizes cache utilization. Delve into topics such as cache coherence, prefetching, performance optimization, and memory allocation. Learn about various data structures like linked lists, vectors, hashmaps, and binary trees, and their impact on cache performance. Discover techniques for memory alignment, insertion optimization, and the use of performance measurement tools. Understand concepts like false sharing and data locality to write more efficient Rust code.
Caches and You: Optimizing Rust Data Structures for CPU Performance - Rust Cologne
Overview
Syllabus
Intro
Overview
Caches
How Caches Work
Cache Coherence
Cache Size
Prefetch
Performance
Matrix Multiplication
In Memory
Cache Line
Heap Allocation
Object Size
Data locality
Separate vector
Code size matches
What is false sharing
How I benchmark
Dont access memory
Any questions
Data structures
Linked lists
Vectors
Hashmap
Binary Tree
Summary
Memory alignment
Insertion
Performance Measurement Tools
Question
JavaScript Source Map
Taught by
Rust
Related Courses
-
The Performance Price of Dynamic Memory in C++
-
C++ Performance Optimization: Algorithmic Complexity, Data Locality, Parallelism, and Compiler Optimizations
-
What Do You Mean by Cache Friendly - Optimizing Data Structures for CPU Performance
-
Bay Area Rust Meetup - Parallel Programming and Concurrent Data Structures
-
Optimizing Binary Search - Advanced Techniques and Performance Improvements
-
Profiling Code in Rust - Techniques for Time and Memory Optimization
