How does Pandas optimize datetime arithmetic and resampling internally?

Datetime operations use int64 nanosecond timestamps for fast arithmetic and grouping.

Pandas stores datetime values as int64 nanoseconds since epoch. This enables fast arithmetic, comparisons, and resampling. Resampling converts timestamps into fixed-frequency bins using efficient interval mapping algorithms.