Explore the intricacies of Java HotSpot performance optimization in this comprehensive conference talk from Strange Loop. Dive deep into JIT and Runtime optimizations, HotSpot garbage collection, and various garbage collectors available in HotSpot. Learn from Monica Beckwith, an experienced Java Virtual Machine expert, as she shares insights gained from over a decade of optimizing JVM heuristics, improving JIT code quality for different processor architectures, and enhancing garbage collection for server systems. Discover the trifecta of performance: responsiveness, footprint, and throughput, and how they relate to garbage collection, compilation, and runtime. Gain valuable knowledge on topics such as execution engines, heap management, locking improvements, escape analysis, compressed OOPs, and tiered compilation. Understand the complexities of Java object headers, string objects, and the generational Java heap. This talk is essential for Java developers and performance engineers looking to optimize their applications and systems running on the Java Virtual Machine.
Overview
Syllabus
Intro
About Me
Establishing the Characters and The Plot • HotSpot Virtual Machine
The Storyline
Java Runtime Environment
The Helpers
Execution Engine
Heap Management
HotSpot VM Runtime
Runtime Goal
The Trifecta of Performance
Responsiveness and GC
Responsiveness and Compilation
Responsiveness and the Runtime
Uncontended Locks
Locking Improvements
Contended Lock Optimization
Footprint and GC
Footprint and Compilation
Escape Analysis
Footprint and the Runtime
A Java Object Header
Compressed OOPs and Compressed Class Pointers
Throughput and GC
Without Intrinsics
Java String Object
Java String Deduplication (G1)
The Generational Java Heap
Compilation Engine
Tiered Compilation
Dynamic Deoptimization
Taught by
Strange Loop Conference
