AI Mode
genesi-ai-mode
A systemd daemon (genesi-aid) that watches for local AI workloads — Ollama, llama.cpp, vLLM, LocalAI — and retunes the whole machine for inference: CPU governor to performance, transparent huge pages, lower swappiness, raised scheduling priority pinned to physical performance cores, and I/O readahead tuned for large GGUF files. The moment inference stops, every change is reverted so your compiler gets the resources back.
- 15–25% faster CPU inference
- Plasma 6 widget + Monitor app
- Live tokens/s readout
- Fully automatic & reversible