Establish a Z[ζ]-specific lower bound for Clifford+√T synthesis

Establish a Z[ζ]-specific non-Clifford-count lower bound for deterministic, ancilla-free approximate synthesis of general single-qubit unitaries over the Clifford+√T gate set, thereby determining whether the proposed circuits attain the gate set’s information-theoretic optimum.

Background

The paper introduces a direct, deterministic, ancilla-free approximate-synthesis algorithm for general single-qubit unitaries over Clifford+√T. Its search operates over the larger cyclotomic ring Z[ζ], where ζ is a sixteenth root of unity, and empirically produces circuits with lower resource-state cost than the provably T-count-optimal Clifford+T construction.

For Clifford+T, an existing lower bound over Z[ω] establishes optimality of the corresponding synthesis method. The paper does not have an analogous lower bound for Z[ζ], so although the Clifford+√T circuits are empirically cheaper, their optimality is unresolved. A Z[ζ]-specific lower bound would determine whether the observed circuits are merely advantageous or actually optimal under the relevant cost model.

References

We do not establish optimality, as the $Z[\omega]$ lower bound of is not known to transfer to $Z[\zeta]$.

Approximate synthesis of general single-qubit unitaries over the Clifford+$\sqrt{T}$ gate set  (2609.16659 - Weiden et al., 15 Sep 2026) in Section 1, Introduction; Section Discussion, subsection “On optimality” and paragraph “Open directions”