Research Roundup for November 2023

By Dr Chris Mansell, Senior Scientific Writer at Terra Quantum

Shown below are summaries of a few interesting research papers in quantum technology that we have seen over the past month.

Hardware

Title: Electron charge qubit with 0.1 millisecond coherence time
Organizations: Argonne National Laboratory; University of Chicago; Lawrence Berkeley National Laboratory; The NSF AI Institute for Artificial Intelligence and Fundamental Interactions; Massachusetts Institute of Technology; Northeastern University; Stanford University; University of Notre Dame
Electron charge qubits on solid neon are not as well known as some of the other quantum computing hardware platforms, but they soon could be. The authors of this paper have increased the relaxation time of the qubits by a factor of 10 and their coherence time by a factor of 1000. They also achieved a readout fidelity of 97.5% and a one-qubit gate fidelity of 99.95%. These metrics are extremely impressive by the standards of any type of quantum processor. Of course, to properly establish solid neon as one of the leading systems, two-qubit gates will have to be demonstrated. Indeed, the paper includes experimental results demonstrating the strong coupling of qubits with the same resonator, which is a good first step in this direction.
Link: https://www.nature.com/articles/s41567-023-02247-5 

Title: Hotter is Easier: Unexpected Temperature Dependence of Spin Qubit Frequencies
Organizations: Delft University of Technology; Netherlands Organization for Applied Scientific Research
Semiconductor spin qubits are subject to a poorly understood effect whereby microwaves cause the spin Lamor frequency to transiently shift. This degrades the fidelity of microwave-driven operations, such as resonant Rabi oscillations. The existing methods to mitigate this effect all have considerable drawbacks, so gaining a better understanding of the phenomenon would improve the prospects of employing these qubits in a scalable quantum processor. In this paper, careful experiments were performed over a range of qubit temperatures. These provided some insights into the underlying microscopic mechanisms that might be at play. Most notably, the frequency shift was smaller at higher temperatures. This is a nice result because it eases the cooling requirements for the qubits.
Link: https://journals.aps.org/prx/abstract/10.1103/PhysRevX.13.041015 

Title: An integrated atom array — nanophotonic chip platform with background-free imaging
Organizations: University of Chicago; Argonne National Laboratory
Most of the equipment in a cold atom lab is big and bulky, which works well for trying out new ideas. However, it is not great for reliability. Making the set-up a lot more compact has been an important goal of the field for decades. In this paper, impressive feats were performed on a millimetre-scale silicon nitride on silicon chip. Single caesium atoms were loaded into a defect-free, 8-by-8 array of optical tweezers. They were imaged with a new, high-fidelity, background-free detection scheme. The chip consisted of nanophotonic cavities embedded in waveguides that took up only 1% of the chip’s surface. This presents an opportunity to integrate many other components directly onto the chip.
Link: https://arxiv.org/abs/2311.02153 

Title: Quantum error mitigation in quantum annealing
Organizations: D-Wave Systems; Simon Fraser University; The University of British Columbia
A beautiful aspect of the quantum technology field is that extremely diverse systems and methods are unified by both quantum theory and information theory. This means that tools and techniques developed in one scenario can be adapted to many others. In this instance, quantum error mitigation protocols devised for the circuit model of quantum computation have been repurposed for quantum annealing. Zero-Noise Extrapolation, implemented on a prototype D-wave Advantage2 processor with 232 superconducting qubits, effectively extended the coherent annealing time by almost an order of magnitude. To put this in perspective, in the past two decades, the coherence time of state-of-the-art superconducting qubits has been improving at a rate of approximately one order of magnitude every four years. 
Link: https://arxiv.org/abs/2311.01306 

Title: Benchmarking Quantum Processor Performance at Scale
Organization: IBM
Quantum volume is one of the most popular measures of quantum processor performance. It has several nice features, including putting an equal emphasis on circuit width and circuit depth. However, it also has some drawbacks, so this paper introduces a complementary measure called layer fidelity. Its aim is to provide insights into the ability of a processor to run rectangular circuits (e.g., ones where the width is greater than the depth), especially those where the logic gates are organised into layered patterns. The layer fidelity was determined for two IBM devices: the 127-qubit Eagle processor and the 133-qubit Heron processor. The latter had a better “error per layered gate” value because the coupling between its adjacent qubits could be switched on and off as required.
Link: https://arxiv.org/abs/2311.05933 

Software

Title: Absence of barren plateaus in finite local-depth circuits with long-range entanglement
Organizations: Tsinghua University; Tencent Quantum Laboratory
Quantum circuits are often designed to depend upon numerical parameters that can be iteratively updated in order to minimise the value of a loss function. Unfortunately, so-called barren plateaus can appear in the loss landscape of deep circuits, making it extremely challenging to decide how to update the parameters. This has been observed to occur when the circuit expressivity is too high, when the initial quantum states are too entangled, when every qubit in the circuit is measured, and when there is too much noise. In this preprint, a distinction is made between the depth of the entire circuit and the local depth, which is the number of non-commuting gates acting on a given qubit. The researchers find that keeping the local depth constant prevents barren plateaus, even when the overall circuit is deep. 
Link: https://arxiv.org/abs/2311.01393 

Title: Hacking Cryptographic Protocols with Advanced Variational Quantum Attacks 
Organizations: Multiverse Computing; University of Navarra; Donostia International Physics Center; Ikerbasque Foundation for Science
Cryptanalysis is the breaking of encryption protocols. As illustrated by Shor’s and Grover’s algorithms, it is an arena in which quantum algorithms can shine. Last year, a variational quantum algorithm was found to be surprisingly good at attacking symmetric cryptography. In a new paper, the algorithm has been made more resource-efficient, now requiring fewer qubits and a shallower circuit. When tested, it cracked the Blowfish encryption protocol with 24 times fewer attempts than a classical brute-force approach. This is an intriguing sign because there aren’t any known classical algorithms that can do better than brute force. In a field as important as this one, when a heuristic method just seems to work, it is certainly worth investigating further. 
Link: https://arxiv.org/abs/2311.02986 

Title: On the Pauli Spectrum of QAC0
Organizations: Columbia University; UC Berkeley
Analysing the Fourier spectra of Boolean functions has produced important results in theoretical computer science. In this preprint on quantum computation, the authors wondered if considering the Pauli spectrum of a many-qubit operator could be equally productive. Instead of looking directly at unitary operations, they distinguished their work from prior approaches by considering quantum channels. This enabled them to prove some new results related to average-case circuit lower bounds and sample-efficient learning algorithms. Overall, their novel way of analysing Pauli spectra seems to allow natural connections to be made to computational complexity. 
Link: https://arxiv.org/abs/2311.09631 

Title: Multi-client distributed blind quantum computation with the Qline architecture
Organizations: Sapienza Università di Roma; Sorbonne Université; VeriQloud; University of Edinburgh; National Quantum Computing Centre
Blind quantum computation (BQC) is a form of secure delegated computation in which a server can run an algorithm for a client without obtaining any information about its input or output. Remarkably, even the algorithm itself can remain hidden from the server. Recently, classical protocols, such as federated machine learning, that involve multiple clients have been gaining prominence. In this paper, an architecture for multi-client, universal BQC was proposed. The clients do not need to possess a great deal of quantum technology, only the ability to perform single-qubit rotations. Due to these fairly simple requirements, the scheme is relatively practical as well as secure. 
Link: https://www.nature.com/articles/s41467-023-43617-0 

November 29, 2023