Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

TT: Fachverband Tiefe Temperaturen

TT 18: Superconductivity: Poster

TT 18.35: Poster

Montag, 18. März 2024, 15:00–18:00, Poster C

Machine Learning assisted Quantum Error Correction on the Rotated Surface Code — •Theo Haas, Kai Meinerz, and Simon Trebst — Theoretical Physics, University of Cologne, Germany

In going beyond the current era of noisy intermediate-scale quantum (NISQ) processors, quantum error correction will be an indispensable tool to reach fault-tolerant quantum computing. However, the originally developed class of combinatorial decoders, such as minimum-weight perfect matching (MWPM) and union find, exhibit scaling behavior that will not allow to decode O(10⁵ - 10⁶) qubits within one clock cycle. Here we discuss refinements to a recently suggested multi-level decoder that combines machine learning and combinatorial decoder in a hierarchical manner. Simulating this 2-stage decoding for the rotated surface code of varying instances, we show that (i) we can efficiently decode O(10⁶) qubits, while (ii) pushing the error threshold beyond the reach of conventional decoders. We further explore the potential of multi-level machine-learning decoders and their Implementation on FPGA platforms. their Implementation on FPGA platforms.

Keywords: Quantum Error Correction; Rotated Surface Code; Machine Learning