A coherence window boundary scan checks whether the total elapsed stage time fits inside a fixed window, with zero margin as the pass/fail edge. The rows use elapsed times 9, 10, and 11 against a 10-unit window, giving margins 1, 0, and -1 (pass, pass, fail) exactly for the stated toy timing ledger; real experiments add clock jitter, decoherence drift, stage calibration, and timing-resolution uncertainty.
Elapsed stage time is compared with the declared coherence window. Exact arithmetic here means exact results for the stated model inputs; measured inputs still carry uncertainty and significant-figure limits.
highlighted = computed this step
Elapsed stage time is compared with the coherence window
The same window is kept fixed; each row changes the declared stage durations.
m=Twindow−Telapsed
Three elapsed times bracket the window edge
The middle row exactly uses the whole window. The last row exceeds it.