Notebookverified

GPU synchronization changed the ranking

Adding synchronization reversed an apparent latency advantage between two inference paths.

benchmark-observationexp-2026-07-18-gpu-03benchmark-protocol

Context

Two inference paths were timed with perf_counter_ns. The supposedly faster path launched more asynchronous work.

Observation

Without device synchronization, the benchmark mostly measured host dispatch. Synchronizing before and after each measured region reversed the ranking.

Questions

  • Should end-to-end latency and device-only latency appear as separate metrics?
  • Which synchronization overhead belongs to the real request path?

Next steps

Repeat with device events, preserve all samples, and document the measured boundary in the result manifest.