Dice Question Streamline Icon: https://streamlinehq.com

Optimal T-count for variable-angle rotations at constant T-depth (catalytic)

Determine whether there exists a catalytic Clifford+T circuit with constant T-depth that, for variable single-qubit z-rotation angles, achieves an epsilon-approximation using T-count O(log(1/epsilon)).

Information Square Streamline Icon: https://streamlinehq.com

Background

The paper shows that single-qubit z-rotations can be implemented with T-depth 3 using a catalyst state. For a fixed rotation angle, both the catalyst size and T-count scale as O(log(1/epsilon)). However, when the rotation angle varies, the presented constructions increase the resource scaling to O(log2(1/epsilon)).

The authors explicitly ask whether the variable-angle case can match the fixed-angle scaling while retaining constant T-depth in the catalytic setting, i.e., whether the T-count can be reduced to O(log(1/epsilon)) for variable-angle rotations.

References

We leave the following questions for future work. First, note that for fixed rotation angle the size of the catalyst state and the $T$-count is $\mathcal{O}(\log (1/\epsilon))$. However, if the rotation angle is variable, these numbers change to $\mathcal{O}(\log2(1/\epsilon))$. Can there be a constant $T$-depth catalytic circuit with $\mathcal{O}(\log (1/\epsilon))$ $T$-count?

Catalytic $z$-rotations in constant $T$-depth (2506.15147 - Kim, 18 Jun 2025) in Section 3 (Discussion)