Secrets of High Performance in Multicore Systems ☠️
summary
Explains when to shard work on atomic values and concurrent maps to improve throughput on multicore systems, and when that is beneficial. Describes algorithms such as biased locking for reader-writer locks and the technique of optimistic reads, and when each should be used to boost performance.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:13Atomic operations and MESI
- 17:09Single atomic counter problem
- 23:15Sharding atomic counters
- 27:57Local cache and R/W load
- 33:00Optimistic reads and versioned lock
- 38:03Bravo lock algorithm
- 42:00Summary
- 43:09Q&A
related talks
52mSecrets of high performance in multicore systems ☠️
39mLet's talk concurrency
48m“Is Parallel Programming still hard? PART 2 of 2”
29mConcurrency and Parallelism From The Ground Up
26mTackling Contention: The Monsters Inside the 'sync.Locker'
36mThe FIDS Theorems: Tensions between Multinode and Multicore Performance in Transactional Systems
01h 12mData parallelism from a multicore perspective
59mAn Introduction to Multithreading in C++20