2 ms·

Neat post but would have loved to have seen a side-by-side comparing the suboptimal code with the optimal code.