Berlin 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
QI: Fachverband Quanteninformation
QI 29: Quantum Information: Concept and Methods II
QI 29.9: Talk
Thursday, March 21, 2024, 17:15–17:30, HFT-TA 441
Hardware-Tailored Mutually Unbiased Bases — •Kyano Levi1,2, Eric Kuehnke2, Daniel Miller2, and Jens Eisert2 — 1Technische Universität Berlin — 2Freie Universität Berlin
Gathering sufficient statistics for quantum state tomography poses a bottleneck for applications on current quantum hardware. In our work, we address this bottleneck by constructing hardware-tailored and near-optimal quantum circuits for diagonalizing entire stabilizer groups on n≤ 6 qubits.
We use the new circuits to craft sets of mutually unbiased bases that are tailored to important hardware connectivities. Our results allow for an exponential reduction of the sample complexity when performing full-state tomography on up to n = 12 qubits, as verified by theory, simulations, and experiments on quantum hardware. Specifically, we achieve a 1.8x shot reduction over state-of-the-art tensor product bases for three-qubit full-state tomography with a cloud-based quantum computer. With classical simulations, we demonstrate a theoretical shot reduction of 11.4x for n = 11 qubits.
Our circuits feature a two-qubit gate depth that is within reach for present-day, erroneous quantum hardware, rendering them a versatile building block for meaningful applications on existing devices.
Keywords: quantum state tomography; circuit optimization; stabilizer circuits; mutually unbiased bases; graph states