Performance Benchmarking
Measure and track graph intelligence performance
Establish baselines and identify bottlenecks.
Benchmarking Script
Expected Performance
Environment: 16 cores, 64GB RAM
| Algorithm | 1M edges | 10M edges | 100M edges |
|---|---|---|---|
| PageRank (20 iter) | 0.5s | 5s | 50s |
| Louvain | 0.8s | 8s | 90s |
| Components | 0.1s | 1s | 10s |
| Triangle Count | 1s | 15s | 300s |
Continuous Benchmarking
Track performance over time:
Best Practices
- Consistent hardware: Same env for comparisons
- Warm-up runs: Discard first run (cold start)
- Multiple iterations: Average over 3-5 runs
- Version tracking: Pin Raphtory version
See Also
- Optimization Guide - Performance tuning
- Resource Limits