An Introduction to JVM Performance

on YouTube ↗59m

" summary

Explains how HotSpot turns bytecode into machine code via levels of JIT profiling and monomorphic call sites so that tiny changes (one extra runtime type) can break inlining and slow a program hugely — example: rewriting a batch to use fewer runtime types drops runtime from ten hours to four minutes. Covers escape analysis, branch-prediction effects, and why logging and megamorphic utilities kill optimizations. Shows practical tools and flags: JMH for benchmarks, JITWatch and async-profiler/Flight Recorder plus -XX compilation prints.

AI-written summary · about

" chapters (19)

1. 0:00Introduction
2. 3:00Performance disclaimer
3. 5:00The basics
4. 7:10The life cycle
5. 15:22The most crucial optimization
6. 17:30Why call a method expensive
7. 21:28Speculative Optimization
8. 23:05Linked Calls
9. 24:07Virtual Method Process
10. 27:04Inlining
11. 28:52Avoiding types
12. 32:06Example
13. 32:45Alternative
14. 34:15Lock Methods
15. 39:56JavaScript
16. 44:06Ifelse
17. 49:08Object Allocation
18. 52:36Benchmarks
19. 56:51Conclusion

" related talks

• 41mAn Introduction to JVM PerformanceRafael Winterhalter · GOTO Oslo 2020
• 01h 01mAn Introduction to JVM PerformanceRafael Winterhalter · JavaOne 2015
• 49mAn Introduction to JVM PerformanceRafael Winterhalter · Devoxx Belgium 2017
• 01h 09mWhat it means to be Just In TimeAshley Puls · SpringOne2GX 2015
• 01h 09mFrom Bytecode to Native Code in HotSpot JVMIonut Balosin · Voxxed Days Bucharest 2016
• 54mJava performance techniques The cost of HotSpot runtime optimizationsIonut Balosin · Voxxed Days Vienna 2017
• 55mThe Performance Engineer's Guide To Java HotSpotMonica Beckwith · Strange Loop 2017
• 35mJava at Speed: Building a Better JVMSimon Ritter · JCON 2020