Requests-per-second climbing through the warm-up phase of a load test for a cached versus baseline API deployment, showing how fast each reaches steady-state throughput as concurrency ramps.