2026 article
QFlux: Quantum Circuit Implementations of Molecular Dynamics. Part III - State Initialization and Unitary Decomposition
Soudackov, A. V., Cabral, D. G. A., Allen, B. C., Dan, X., Vu, N. P., Cianci, C., … Batista, V. S. (2026, January 30). (Vol. 1). Vol. 1.
open access via doi.orgThis tutorial builds upon the foundations established in Part~II to present a unified, implementation-oriented overview of quantum state initialization and unitary decomposition for n-qubit systems, available in QFlux. We begin with the preparation of arbitrary quantum states and develop two complementary constructions: (i) a recursive multiplexor method (after Shende et al.) that disentangles qubits via multiplexed Ry and Rz rotations, and (ii) an algebraic scheme based on uniformly controlled rotations (UCRs) (after Möttönen et al.) that realizes the same mapping through analytically defined rotation networks with predictable gate counts. We then extend these state-preparation tools to generic unitary synthesis through Givens-rotation, column-by-column, and recursive cosine-sine (CSD) and quantum Shannon (QSD) decompositions, explicitly linking linear-algebraic factorizations to executable circuits while tracking CNOT complexity. Finally, we introduce the Walsh decomposition for diagonal unitaries and show how Gray-code ordering and local CNOT cancellations yield an \(O(2^n)\) entangling-gate cost, providing shallow, NISQ-friendly implementations. Taken together, these techniques form a pedagogical bridge from matrix analysis to hardware-efficient circuit constructions, offering clear design rules, closed-form parameters, and scalable synthesis pathways for simulation and experiment.