Making machine
analysis clear.

Possibility fractures. Each point is one of 64 assignments to six Boolean variables; edges join assignments differing in one bit. Four clauses enforce equality of the first three bits. The 16 surviving assignments form two separate cubes. This finite example illustrates a solution landscape; it proves no universal SAT lower bound. For research on clustering in random SAT, see Mézard, Mora & Zecchina.

Hidden structure. Three outer products a ⊗ b ⊗ c combine into a 4 × 4 × 4 tensor. Each contribution is rank one; their sum reconstructs the array. This supplied decomposition is not a proof of minimum tensor rank. See Kolda & Bader’s tensor decomposition survey.

A limit remains. The heights show real amplitudes for Grover search with one marked item among N = 64, starting uniformly. The oracle flips its phase; reflection about the mean amplifies it. Measurement probabilities are squared amplitudes. The film shows six iterations. Grover’s algorithm takes O(√N) queries; the matching Ω(√N) lower bound concerns unstructured black-box search with constant success probability, not every quantum task.

Motion between quantum steps is a geometric interpolation; the held endpoints are the calculated oracle and diffusion states. Ω is an asymptotic lower bound, not “best case” or necessarily a tight bound. Only the final chapter makes a lower-bound claim; each mathematical model has its own assumptions.