JavaScript JITs
summary
Live demos show assembly from SpiderMonkey, JaegerMonkey, IonMonkey and Crankshaft and claim real JITs hit 100 million iterations per second on a numeric loop, get property access down with inline caching, and handle tricky semantics like overflow checks and negative zero handling. A type-specializing JIT reduces instruction counts and runs within 2–3x of C on benchmarks; prototypes, GC and shape changes require cache invalidation but the patched fast-paths work and speed up real workloads.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 10:44Objects
- 17:40Type Compiler
- 19:06ESI
- 20:34Performance
- 21:04Semantic Gap
- 23:21Flexure
- 24:08Lineup
- 24:28Suggestions
related talks
25mJavaScript engines how do they even?
31mJavaScript Compilation Techniques or "Wasn't JavaScript supposed to be slow?"
17mDemystifying (JavaScript) engines
10mTypes, TypeScript and the JIT
18mDemystifying JS Engines
36mFast and Dynamic
23mJavaScript Engines: The Good Parts™
29mJavaScript in JavaScript: Inception