C++ Performance Lab

Measure sequential and parallel computation using a real C++ program. Run a single dataset or compare performance across five sizes.

Median of seven timed runs per mode after two warm-up runs. Dataset generation is excluded. Parallel timings include thread creation and joining. Speedup = sequential time / parallel time; below 1× means parallel was slower. Measurements are affected by CPU, compiler, load, and scheduling. This is an educational demonstration, not a controlled hardware benchmark.