The Static Dynamic JVM – A Many Layered Dive

on YouTube ↗01h 01m

" summary

Claims that the JVM reduces runtime chaos by turning observed behavior into static decisions and shows concrete mechanisms: Leiden AOT cache runs bootstrap methods early so many lambda call sites become AOT-resolved, improving startup and peak; invoke dynamic bootstrap methods are analyzed and shifted between AOT and JIT; a proposed parametric JVM encodes per-call-site static knobs for reified generics and type classes; stable/mutable call sites and stable variables drive speculations with deoptimization fallback; notes Rice/halting limits (3x+1 tested up to 2^70) and busy‑beaver results up to 5 states.

AI-written summary · about

" chapters (8)

1. 0:00Introduction and concepts
2. 2:25Etymology and definitions
3. 12:30Bridging static and dynamic
4. 13:11Static analysis and modeling
5. 21:14Complexity and the abyss
6. 22:43JVM implementation cheats
7. 52:21Language and VM evolution
8. 58:20Closing thoughts and Q&A

" related talks

• 41mAn Introduction to JVM PerformanceRafael Winterhalter · GOTO Oslo 2020
• 15mThe Hidden Dynamic Life of JavaNataliia Dziubenko · J-Fall 2023
• 18mThe Hidden Dynamic Life of JavaNataliia Dziubenko · Voxxed Days Zürich 2024
• 55mJIT and AOT in the JVMMark Stoodley · JVM Language Summit 2019
• 59mAn Introduction to JVM PerformanceRafael Winterhalter · JavaDay Lviv 2016
• 01h 01mAn Introduction to JVM PerformanceRafael Winterhalter · JavaOne 2015
• 49mAn Introduction to JVM PerformanceRafael Winterhalter · Devoxx Belgium 2017
• 55mMaking Java More DynamicRafael Winterhalter · W-JAX 2018