The Windy Road From Research to Real-World Impact
summary
Project starts from a 2011 vision 1 VM to rule them all and invents a practical partial‑evaluation pipeline (the first Futamura projection at industrial scale) so a single master‑student wrote a JavaScript engine in months by implementing only an interpreter. GraalVM ships inside Oracle Database 23c, supports Python/JS/Wasmtime, faces 1,281‑entry Python C API compatibility pain, and enables ahead‑of‑time native images that give about 30x faster startup. Codebase is ~4M LOC and cloud vendors provide Graal‑compatible APIs.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:06Starting GraalVM
- 3:34Early research in Zurich
- 6:47Partial evaluation
- 10:23Polyglot vision
- 15:18Throwing ideas at the community wall
- 17:26Accidental ahead-of-time compilation
- 20:59Advantages of ahead-of-time compilation
- 24:52Ecosystem adoption
- 26:21Real-world impact
- 28:29Open source
related talks
17mGraalVM Runtime Efficiency and Contributing to the Development Community
20mGraalVM: Run Programs Faster Anywhere
33mNew Opportunities for Developers with GraalVM
03h 07mEverything you need to know about GraalVM
47mJava and the GraalVM
50mOne VM to Rule Them All
59mGraalVM: Run Programs Faster Anywhere
47mPolyglot Done Right with GraalVM