Amdahl's Law: A 2D Scheduling & Speedup Model

This is the 2D companion to the 3D Amdahl's Law cluster scene, reaching the same extended-Amdahl physics — T(N) = (1−P)·T(1) + P·T(1)/N + O·N·T(1) — through a genuinely different visual mechanism: a Gantt-style scheduler diagram instead of a grid of cubes. Every core gets its own lane; the serial slice is drawn only on core 0's lane while every other lane sits idle, the parallel slice lights up all N lanes at once, and the coordination-overhead slice — the synchronization tax that grows with core count even as useful work per core shrinks — closes out the cycle. Alongside the schedule, a directly-computed speedup-vs-core-count curve is plotted for the full N=1..64 range at the current settings, with the closed-form optimal core count N* = √(P/O) marked on it, so you can see not just how fast the current configuration runs but where the whole curve peaks and starts falling as coordination cost overtakes the benefit of more workers.