Performance Documentation¶
Performance analysis, benchmarking, and system optimization guides for ShadowHound.
Overview¶
This directory tracks performance characteristics, bottleneck analysis, and optimization strategies for the ShadowHound system.
Documents¶
- Performance Analysis Plan - Comprehensive performance testing strategy
Key Metrics¶
LLM Backend Performance¶
- Response latency (vLLM vs Ollama vs cloud)
- Token throughput
- Memory usage
- See LLM Integration Hub
Robot Control Performance¶
- Command execution latency
- Skill execution time
- ROS2 topic rates
- WebRTC stream quality
System Resources¶
- CPU utilization (laptop, Thor, GO2)
- Memory consumption
- Network bandwidth
- Power consumption
Validation¶
- [ ] Baseline performance metrics documented
- [ ] Bottlenecks identified and tracked
- [ ] Optimization efforts measured and validated
See Also¶
- LLM Integration - Backend performance tuning
- Troubleshooting - Performance debugging
- Software Hub - Software stack optimization