Wallraff, Andreas

Date:   Thursday, July 30, 2020
Time:   10:00
Place:   scheduled Zoom meeting
Host:    Klaus Ensslin

Running Optimization Algorithms on a Small Superconducting Quantum Processor

Andreas Wallraff
ETH Zurich

Superconducting circuits are a prime contender both for realizing fault-tolerant quantum processors and for executing noisy intermediate scale quantum (NISQ) algorithms. After a brief introduction to the quantum physics of superconducting circuits, I will discuss our implementation of quantum approximate optimization algorithms (QAOA), one of a whole class of variational algorithms which are believed to be promising to solve computationally hard problems on quantum computers. Such algorithms also emerge as a benchmark for the performance of NISQ-era quantum information processing systems. Gaining computational power from QAOA critically relies on the mitigation of errors during the execution of the algorithm, which for coherence-limited operations is achievable by reducing the gate count. Here, we present an improvement of up to a factor of three in algorithmic performance, by implementing a continuous hardware-efficient gate set. This gate set allows us to perform a critical step in QAOA with a single operation on each pair of qubits instead of decomposing it into multiple gates. We experimentally investigate the circuit-depth-dependent performance of our QAOA implementations for finding solutions to exact-cover problems mapped onto three or seven qubits using up to a total of 399 operations in up to 9 layers. Our results demonstrate that the use of continuous gate sets may be a key component in extending the impact of near-term quantum computers [1]. In the same device architecture, we have recently demonstrated repeated error detection [2], an important step toward quantum error correction and the execution of fault-tolerant quantum algorithms.

References
[1] N. Lacroix et al., arXiv:2005.05275 (2020)
[2] C. K. Andersen et al., Nat. Phys. (2020). https://doi.org/10.1038/s41567-020-0920-y
 

JavaScript has been disabled in your browser