arXiv:2610.28057
Optimal Quantum-Circuit Synthesis and Implementation of General Unitaries with Parallelized Bitstring-Conditioned Rotations
Heikki Suominen, Suman Kundu, Heidi Kivijärvi, Sasu Tuohino, Mikko Tuokkola, Eemeli Forsbom, Madhavi Chand, Vasilii Vadimov, R. Vijay, Mikko Möttönen
Abstract
We theoretically introduce a quantum-circuit synthesis protocol to realize any $n$-qubit transformation $U\in U(2^n)$, up to a diagonal. The resulting circuit is directly native to and optimal in gate count and depth for systems where bitstring-conditioned rotations with tunable angle and axis, $C^{(n-1)}R(θ,φ)$, form the set of native operations. Subsequently, we suggest how this protocol can be implemented in a few-mode superconducting circuit, and experimentally demonstrate the method for the smallest nontrivial case, i.e., for two qubits. We report an average gate fidelity of 97.6 % for Haar-random two-qubit unitaries up to phase shifts in the computational basis. Importantly, these results suggest that multimode devices, such as those in the multimon or multiunimon families may become competitive with qubit-array networks for general-unitary implementations for qubit counts as low as three.
AI Summary
The paper studies synthesis of dense $n$-qubit unitaries for hardware with native bitstring-conditioned rotations $C^{(n-1)}R(\theta,\phi)$, assuming rotations on disjoint computational-basis transitions can be driven simultaneously with independently tunable amplitudes and phases. A modification of cosine-sine decomposition implements any $U\in U(2^n)$ up to a final diagonal using $2^n-1$ layers, each containing $2^{n-1}$ parallel rotations. Diagonal phases are propagated through the decomposition, converting uniformly controlled rotations into rotations about independently chosen axes in the $xy$ plane. A hypercube matching argument bounds parallelism, and parameter counting establishes gate-count and depth optimality within this native-operation model. A variant uses only $\pi/2$ rotations, requiring $2^{n+1}-2$ layers; another procedure wraps rotation angles into $(-\pi,\pi]$. The experimental demonstration uses a two-qubit subspace of a superconducting trimon. Haar-random $4\times4$ transformations compile into three layers of 64-ns multiplexed pulses. Randomized benchmarking, with residual diagonals absorbed into subsequent transformations, gives a best-decile fidelity of $97.6\pm0.2\%$. Control includes calibration and dynamic compensation of drive-induced frequency shifts. The final diagonal is dispensable before computational-basis measurement, but must otherwise be retained or absorbed into later operations. Short coherence times and overlapping transitions prevented a three-qubit demonstration. Larger-system fidelity comparisons are estimates rather than measurements. Frequency-shift calibration scales exponentially: approximately four hours for the demonstrated two-qubit system and an estimated 36 hours for three qubits. The depth advantage assumes independently addressable, simultaneously driven transitions.
Error found: The abstract omits that the reported 97.6% average gate fidelity is a best-decile result selected from a time series of randomized-benchmarking fidelity estimates, rather than an unselected estimate of typical performance.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 15 pages, 7 figures, 1 table