Benchmarks¶
Loading benchmark data...
Results at a glance¶
Iteration micros (µs/op)¶
Scenarios (µs/op)¶
Relations (µs/op)¶
Unified delta (µs/op)¶
Allocation per tick (B/op, japes only)¶
Detailed benchmark pages¶
| Page | What it measures |
|---|---|
| Methodology | Hardware, JMH config, the @Setup / @Benchmark contract |
| Iteration micros | Pure per-entity read/write cost at 1k / 10k / 100k |
| N-Body | One integrator system, Euler step, 1k + 10k bodies |
| Particle scenario | Five-system game loop with 1% per-tick turnover |
| Sparse delta | Change-detection: 100 dirty per tick, 1 observer |
| Realistic tick | Three observers, 10k/100k, st/mt |
| Predator / prey | Relations: @Pair vs @ForEachPair vs Bevy |
| Unified delta | Multi-target @Filter vs Zay-ES EntitySet |
| Valhalla EA | JDK 27 JEP 401 value records |