Core metric families

• Model quality: loss, task metric, calibration error, drift score.

• Training efficiency: convergence rate, compute cost, scaling efficiency.

• Hardware efficiency: utilization, bandwidth use, latency, perf-per-watt.

• Workflow impact: runtime savings, signoff-correlation quality, false recommendation rate.

• Deployment safety: compatibility pass rate, rollback readiness, incident SLA.

BEFORE / AFTER — AI-VLSI Closure

metric quality
  ^
  |                        o target region
  |                 o post-fix + regression
  |            o
  |      o baseline failing run
  +-----------------------------------------> iteration
      evidence audit   fix    full validation