Option 2: Theory of Non-Markovian Noise Correction in Multi-Qubit Gate Operations
Abstract
This theoretical work will produce dynamically corrected gates specific to overcoming the barrier posed by non-Markovian noise to dependable high-fidelity operations in silicon qubits. Recent experiments by Profs. Dzurak and Morello at the University of New South Wales (UNSW) indicate that average single-qubit gate fidelities are currently held down by coherent errors stemming from highly non-Markovian noise sources (primarily low-frequency drift in the qubit resonance frequency) in spin qubits in both Si-MOS quantum dots and Si:P donors. Likewise, two-qubit operations (which have been realized but not yet as carefully characterized) will be even more severely affected by low frequency noise due to the significantly longer gate times involved. While the resulting coherent errors typically cannot be treated by quantum error correction protocols, they are naturally treatable by using pulse shaping or compensating pulse sequences. This project aims to produce ten-fold improvement in both the single-qubit and two-qubit gate fidelities in a multi-qubit Si-based system by combining the pulse shaping and pulse sequence approaches. This approach has great promise as a breakthrough in high-fidelity quantum control that could be applied across all Si spin qubit platforms. Single-qubit fidelities will be raised without sacrificing much speed by means of analytical pulse shaping via partial-reverse engineering. This novel approach not only allows optimization for gate time, realistic pulse form, and dynamical robustness, but also can be used to reduce leakage outside the logical subspace during gating in both single- and multi-qubit settings, all without getting lost in the large control landscape. Two-qubit fidelities will be raised by making use of the typically higher relative precision of the single-qubit gates (even before pulse shaping, but especially after) to bootstrap the fidelity via novel echo sequences. In principle , all two-qubit coherent error can be corrected to arbitrary order in the quasistatic limit by concatenating sequences, as long as one has access to sufficiently precise single-qubit gates. As the method is completely modular, it can be applied regardless of the technique used to implement the single-qubit gates and the interqubit coupling. Generality of the approach notwithstanding, this work will be specifically carried out in collaboration with the ongoing experimental efforts at UNSW in order to produce realistic qubit control protocols tailored to their experiments that will actually demonstrate marked advantage in the lab.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 11, 2018
- Source ID
- W911NF1710287
Entities
People
- Jason Kestner
Organizations
- Army Contracting Command
- National Security Agency
- University of Maryland, Baltimore County