Fingers involve splitting a single large transistor into multiple smaller transistors (or “fingers”) and connecting them in parallel.
Instead of having one large transistor, the width is divided into several parts. Each finger shares a common gate but has separate source and drain regions.