Orlix Native Performance
Performance claims need workload evidence, not architecture intent.
Required Evidence
- Exact binary or workload.
- Exact simulator or device target.
- Build configuration.
- Trigger command or user action.
- Baseline or previous run.
- Measurement command or profiler capture.
- Hotspot, latency, throughput, or allocation result.
- What was not measured.
Initial Orlix Focus
- ELF launch latency.
execveand dynamic-loader path cost.- Syscall throughput for hot paths.
- PTY read/write throughput and readiness latency.
- Rootfs tar and OCI layout import time.
- Virtio-blk persistence and I/O throughput.
- Virtio-fs path lookup and file operation latency.
- Imported-binary startup latency in iOS Simulator.
Rules
- Do not claim native performance from “compiled successfully.”
- Do not claim full performance from one microbenchmark.
- Do not compare against macOS runtime behavior for Orlix’s initial runtime target.
- Label Simulator-only results as Simulator-only.