Research Summary for April 2022

By Dr. Chris Mansell

Material

Title: Two-Qubit Silicon Quantum Processor with Greater Than 99% Operation Fidelity
Organizations: Princeton; NIST/University of Maryland; Sandia National Laboratories

In January of this year, very high logic gate fidelities for silicon qubits were reported in Nature. However, quantum computers must also have high preparation and state measurement (SPAM) fidelities. In silicon devices, SPAM fidelities varied from around 80-90%, so it was imperative to improve this value. In April, new experimental work pushed it to over 97%. The logic gate fidelities, in ascending order, were equally impressive: 99% for single-qubit gates operating in parallel, 99.8% for a two-qubit phase-controlled gate, and 99.9% when controlling a single qubit. one qubit at a time. Given these numbers and some recent advances in manufacturing methods, the paper, published in Science, argues that silicon spin qubits could evolve into a “dominant” technology for NISQ processors.
Link: https://www.science.org/doi/10.1126/sciadv.abn5130

Title: Experimental Photonic Quantum Memristor
Organizations: University of Vienna; Politecnico di Milano; Consiglio Nazionale delle Ricerche

A memristor is a nanoscale electronic component with the special property that its resistance depends on the amount of charge that has passed through it. Classical memristors are considered extremely attractive candidates for next-generation computing systems. The quantum memristor proposed and demonstrated in this article obeys the same equations as its classical counterpart but it can operate coherently on quantum states. In the device, the photons travel different paths from an integrated photonic chip. When measured, an active feedback loop is used to adjust the reflectivity of the chip’s beamsplitters. The provided nonlinearity and short-term memory were then used in an image classification task. When single photons encoded the data, the experiment was more accurate and more resource efficient than what has been reported for classical memristors. These are encouraging results and it will be interesting to see how future larger scale versions of this setup perform.
Link: https://www.nature.com/articles/s41566-022-00973-5

Title: High coherence and low crosstalk in a tileable 3D superconducting integrated circuit architecture
Organizations: University of Oxford; University of Southampton

This paper reports on a two-dimensional tile containing four superconducting transmon qubits and shows that its design means that many tiles can be placed in a large grid without degrading performance metrics. First, the wires are placed on the opposite side of the tile to the qubits and they come out of the plane so they can avoid each other. Second, the tile is enclosed in a cavity that provides a qubit-specific electromagnetic environment. Single-qubit gate fidelities were found to be 99.982% whether gates were performed sequentially or simultaneously. This implies that the crosstalk is extremely low and helps demonstrate that the tile is scalable. Several calculations in the Supplementary Information indicate that when more tiles are added and the qubits are wired together to allow two-qubit gates to run, the coherence times and fidelities should remain constant.
Link: https://www.science.org/doi/10.1126/sciadv.abl6698

Title: Quantum state preparation and tomography of entangled mechanical resonators
Organization: Stanford University

A quantum acoustic processor with coupling between a superconducting transmon qubit and a pair of piezoelectric, nanomechanical, and phononic crystal resonators might sound like something out of a science fiction novel. On the contrary, it is an experimentally demonstrated device that has rapidly improved in recent years. In this research, improved fabrication methods extended the coherence times of the transmon qubit and resonators. A deterministic iSWAP operation was implemented in approximately 25 nanoseconds with an estimated fidelity of 95% and non-demolition quantum measurements were performed for readout. The plan is to further extend coherence times before incorporating features like quantum random-access memories or Kerr-cat qubits into the architecture.
Link: https://www.nature.com/articles/s41586-022-04500-y

Software

Title: An Analytical Theory of the Dynamics of Extended Quantum Neural Networks
Organizations: The University of Chicago; Chicago Quantum Exchange; IBM; University of Maryland

This paper examines a variational quantum algorithm performing supervised learning and examines how its residual training error (RTE) varies as it is trained by gradient descent. In order to have a chance of arriving at an analytical result, the authors had to consider overparameterized regimes, where the number of trainable parameters is large. They used the theory of neural tangent nuclei to discover that the RTE could converge exponentially. These results were of immediate practical interest as they shed light on the results of some previously published numerical observations. Additionally, the authors were able to identify the types of quantum circuits that could have these desirable, exponentially converging drive dynamics without an unmanageable amount of parameters. In the future, they plan to explore different circuit designs and see if there are any links to the ability of learning algorithms to generalize when given new data.
Link: https://arxiv.org/abs/2203.16711

Title: ADAPT-VQE is insensitive to landscapes with rough parameters and arid plateaus
Organization: Virginia Tech

The variational quantum eigensolver, or VQE, is an algorithm for calculating the energy levels of molecules. This article focuses on an adaptive and problem-tailored version of the protocol known as ADAPT-VQE, where instead of optimizing the variational parameters of the circuit, the circuit structure itself is improved by iterative way. Numerical simulations of several molecules showed that ADAPT-VQE’s initial assumptions helped it avoid local minima. Even if the algorithm finds itself in an awkward parameter regime, such as a local minimum or a sterile plateau, it can “bury itself” towards the exact solution. Some challenges remained, particularly one the researchers called “gradient trough,” so the immediate plan is to investigate them further.
Link: https://arxiv.org/abs/2204.07179

Title: Non-distribution generalization for learning quantum dynamics
Organizations: Technical University of Munich; Munich Center for Quantum Science and Technology; Freie University in Berlin; Caltech; Los Alamos National Laboratory; University of Southern California; ADMIRATION
Quantum Neural Networks (QNNs) continue to be a popular research topic with a new preprint providing the first example of a very useful phenomenon. Before starting a machine learning project, one typically accesses a dataset and splits it into two parts, one to train the model and one to test it. After training, if the model scores high on the test data, we say it showed generalization in the distribution. Nevertheless, the model can be fragile and obtain low scores when tested on alternative or modified data. In this case, we say that it failed to generalize out of the distribution. In the latter work, a QNN learns how unitary operations act on separable states, and then is asked how they would operate on entangled states. The analytical and numerical results showed that the QNN generalized very well from the training distribution to the test distribution.
Link: https://arxiv.org/abs/2204.10268

Title: Quantum self-supervised learning
Organizations: University of Oxford; University of Cambridge

Machine learning (ML) has been incredibly successful over the past decade, so the current trend in quantum computing is to see what ML ideas could be carried into the quantum realm. Recently, self-supervised learning has become an important way to train ML models without the need for human-annotated data. While models that monitor themselves using unlabeled data sound great, the downside is that they are extremely computationally expensive. The idea expressed in this paper is that since quantum neural networks (QNNs) operate in high-dimensional Hilbert spaces, they may be able to learn the complex patterns needed for self-supervised learning. The researchers simulated a QNN without any noise or decoherence and showed that it could help an ML model achieve higher self-supervised accuracy than a purely classical model. When they ran the QNN on IBM’s “Paris” quantum computer, the accuracy matched that of the classical case, showing that it was robust to the imperfections of a NISQ device.
Link: https://iopscience.iop.org/article/10.1088/2058-9565/ac6825

April 27, 2022

Sherry J. Basler