Turning GraalVM from Research to Product
summary
GraalVM moves from research to product by shipping a working runtime: a native-image ahead-of-time compilation path that yields ~40% lower memory, slightly higher peak throughput and much faster startup versus JIT, and production deployments inside Oracle Database and MySQL and at Netsuite. Key engineering choices include practical partial evaluation of interpreters to support polyglot engines, profile‑guided AOT for predictable performance, a decade of incubation, and tradeoffs that turned prototypes into tools that customers actually use.
AI-written summary · about
chapters (14)
- 0:00Research to GraalVM vision
- 3:18GraalVM naming and origins
- 5:47Team scaling challenges
- 7:56Oracle’s GraalVM investment
- 8:39Innovation under constraints
- 12:50Early GraalVM experiments
- 16:26GraalVM and Oracle Database
- 17:32Polyglot programming challenges
- 20:44JNI and language specs
- 24:16Why ahead-of-time compilation outperforms JIT
- 27:03Java runtime trade-offs
- 30:15Operational efficiency impact
- 33:24Ecosystem adoption
- 36:14Closing and future outlook
related talks
03h 07mEverything you need to know about GraalVM
59mGraalVM: Run Programs Faster Anywhere
40mThe need for speed. What is GraalVM and how can it help you?
47mGraalVM for Endless Performance
50mGraalVM for Java developers
52mGraalVM, the polyglot virtual machine
52mGraalVM: Run Programs Faster Anywhere
49mGraalVM: Run Programs Faster Anywhere