Faster Data With Faster Java
summary
Azul’s Platform Prime swaps into existing Java deployments and claims to remove JVM GC pauses using the C4 concurrent compacting collector, replace JIT with the Falcon LLVM-based compiler for higher throughput, and replay warmed profiles with ReadyNow to cut warm-up. No code changes or recompilation are required; examples show 20% higher per-node throughput, a 15% cloud cost reduction for a Kafka user, MasterCard eliminating ~10,000 full GCs/day, and Taboola cutting server footprint >30% and DB heap by ~50%; Linux-only.
AI-written summary · about
chapters (8)
- 0:00Introduction and Azul Systems
- 1:48Azul Platform Core
- 3:15Higher Performance JVM
- 4:40How Azul Improves Performance
- 8:37Java Performance Graphs
- 11:44ReadyNow Technology
- 14:10Customer Case Studies
- 18:43Summary and Q&A
related talks
27mSponsored Session: Cassandra on Steroids
35mJava at Speed: Building a Better JVM
44mJava at Speed: Building a Better JVM.
50mAOT or JIT: Faster Startup or Faster Code?
01h 03mKeeping Your Java Hot.
40mHigh-Performance Java-Cloud-Native Apps & Microservices
51mJava at Speed: Getting the Most Out of Modern Hardware
38mWelcome to the Jungle - A safari through the JVM landscape