Research Roundup for April 2023

By Dr Chris Mansell, Senior Scientific Writer at Terra Quantum

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

Hardware

Title: Experimental Twin-Field Quantum Key Distribution Over 1000 km Fiber Distance
Organizations: University of Science and Technology of China; Chinese Academy of Sciences; Photon Technology (Zhejiang) Co. Ltd.; Yangtze Optical Fibre and Cable Joint Stock Co. Ltd.; Tsinghua University
Quantum key distribution (QKD) enables two remote parties to share a random secret key so they can then communicate privately. QKD’s main practical problem is that the achievable rate at which the key can be generated becomes very low when the communicating parties are separated by long distances. In this experiment, a twin-field QKD protocol was implemented and the key rate was measured for distances from 202 km up to 1002 km. By developing and employing several methods for improving the key rate, the reported results exceeded those of all prior twin-field QKD demonstrations.
Link: https://arxiv.org/abs/2303.15795

Title: Development and Demonstration of an Efficient Readout Error Mitigation Technique for use in NISQ Algorithms
Organization: Rigetti Computing
Approximate state tomography and classical shadow tomography are both methods of estimating observables using efficiently many random measurements. For NISQ devices, measurement errors must be taken into account. The authors of this preprint analysed the impact such errors have on approximate state estimators and considered different strategies to mitigate them. They devised a quick and accurate error mitigation technique that they implemented on Rigetti’s Aspen-11 superconducting quantum processor. They processed one million measurements in less than one and a half minutes, a task that may have taken them seventeen hours without their new approach.
Link: https://arxiv.org/abs/2303.17741

Title: High-Fidelity, Frequency-Flexible Two-Qubit Fluxonium Gates with a Transmon Coupler
Organizations: Massachusetts Institute of Technology; MIT Lincoln Laboratory
Composite superconducting circuit components, such as fluxonium qubits and transmons, have been the subject of considerable research and development. In this preprint, researchers have created a novel superconducting quantum processor based on logic gates between pairs of fluxonium qubits coupled together via a transmon. Compared to prior set-ups, this design straightforwardly produces stronger desired couplings and weaker unwanted couplings. The logic gates, involving microwaves that drive Rabi oscillations, can be tuned to operate over a 2 GHz range, which will enable frequency collisions to be avoided in larger devices with numerous qubits. With state-of-the-art fidelities and coherence times, the researchers already have a number of ideas to further push the forefront of this exciting field. 
Link: https://arxiv.org/abs/2304.06087

Title: Reference-State Error Mitigation: A Strategy for High Accuracy Quantum Computation of Chemistry
Organization: Chalmers University of Technology; 
This paper describes a chemistry-inspired strategy for reference-state error mitigation (REM). Due to the efficient way in which REM can be implemented – with minimal post-processing on a classical computer and either one or no additional measurements – it can be combined with other error mitigation techniques, such as readout mitigation. Small molecules, such as hydrogen and lithium hydride, were analysed using the variational quantum eigensolver algorithm run on two superconducting NISQ processors: ibmq_quito and Särimner. Applying REM improved the computational accuracy of the energy calculations by up to two orders of magnitude.
Link: https://pubs.acs.org/doi/10.1021/acs.jctc.2c00807

Title: Quantum critical dynamics in a 5,000-qubit programmable spin glass
Organizations: D-Wave Quantum Inc.; Boston University; Simon Fraser University
For over 20 years, D-Wave has pioneered the field of quantum annealing. The number of qubits in their devices has been following an impressive upward trajectory. Various scientific investigations have been conducted relating to noise, coherence, entanglement, the types of problems that can be solved and the speed with which they can be solved. Recently, an important issue about the relationship between thermal annealing and coherent quantum annealing was resolved for their 2000-qubit device. Now, building on these results in their new Nature paper, they show coherent dynamics for their larger, more highly connected annealer called “Advantage,” which has 5000 qubits arranged in a “Pegasus” layout. 
Link: https://www.nature.com/articles/s41586-023-05867-2

Software

Title: Robust Dequantization of the Quantum Singular value Transformation and Quantum Machine Learning Algorithms
Organization: Nagoya University
Ewin Tang has shown that most quantum machine learning algorithms in the Quantum Random Access Memory model have no quantum advantage. That is, classical algorithms that can sample quantum states perform equally well. However, François Le Gall has noticed in his new research that it may make more sense to investigate classical algorithms that can only do this sampling approximately. He found that even under this new condition, the classical algorithms still run just as quickly. This result holds for classical versions of six different algorithms, including the quantum singular value transformation algorithm and supervised clustering.
Link: https://arxiv.org/abs/2304.04932

Title: Universal noise-precision relations in variational quantum algorithms
Organizations: Osaka University; JST, PRESTO; RIKEN Center for Quantum Computing
In this paper, the authors analytically estimate the error in the cost functions of variational quantum algorithms subject to Gaussian noise. They give insights into what affects the sensitivity to this noise and show how there is a trade-off relationship between the trainability and the noise resilience of a cost function. They test their estimates by numerically simulating both the variational quantum eigensolver for a spin chain and a toy problem related to variational compiling. They also design a novel error mitigation technique and make their code easily available online.
Link: https://journals.aps.org/prresearch/abstract/10.1103/PhysRevResearch.5.023025

Title: Density-Matrix Renormalization Group Algorithm for Simulating Quantum Circuits with a Finite Fidelity
Organizations: Atos Quantum Laboratory; Université Grenoble Alpes; Cornell University; Flatiron Institute
In this new paper, the researchers ask whether it is harder for a quantum computer to outperform a classical computer for a useful task or a useless one. They point out that useful tasks are related to the real world, which is largely comprehensible and structured. Algorithms, whether classical or quantum, can attempt to take advantage of this structure. Completely useless tasks, on the other hand, don’t have to result in any interesting insights, which means that there is lots of flexibility to design a quantum circuit that is contrived to be extremely hard for a classical computer to simulate. Even though random quantum circuits do just this, the researchers developed a classical algorithm that can simulate them up to hundreds of qubits. Compared to prior classical algorithms, this one has a better scaling with the number of qubits but encounters exponential difficulties when simulating quantum processors with higher fidelities. 
Link: https://journals.aps.org/prxquantum/abstract/10.1103/PRXQuantum.4.020304

Title: A Framework for Demonstrating Practical Quantum Advantage: Racing Quantum against Classical Generative Models
Organizations: Zapata Computing Canada Inc.; Vector Institute, MaRS Centre; University of Waterloo; University of Toronto
The latest generative AI models can look through a large database of images and then generate a new image that is novel, in the sense that it was not in the original database, and yet still has a similar style and the right sort of content. Such models are said to be able to generalise well. Numerically testing this generalisation performance may not seem straightforward. However, the authors of this preprint provide a framework for doing just that. In particular, they show how generative quantum models can be compared with classical ones. They create a synthetic dataset, a proportion of which can be accessed by the models. The newly generated data are compared with the remainder of the original data using a cost function that numerically scores their similarity. The results show that quantum circuit Born machines are very competitive in terms of the diversity of their outputs. Furthermore, they are more efficient than classical models when provided with only limited data.
Link: https://arxiv.org/abs/2303.15626

Title: Quantum-resistance in blockchain networks
Organizations: IDB-Inter-American Development Bank; LACChain-Global Alliance for the Development of the Blockchain Ecosystem in LAC; Quantinuum; Escuela de Ingenieria y Ciencias, Mexico
Current blockchains face critical risks related to their digital signatures and cryptographic keys. Since blockchains are distributed ledgers that store data, the “store now, decrypt later” threat (also known as the “hack today, crack tomorrow” threat) is an especially serious concern – nothing needs to be hacked because it is all already stored. In this paper, the authors address these risks by developing an open-source, end-to-end framework for adapting Ethereum-based blockchains to make them quantum resistant as well as robust and scalable. The framework incorporates quantum sources of entropy, post-quantum certificates, post-quantum keys, the signing of transactions and the on-chain verification of signatures. Three ways of performing this verification step were tested. They involved solidity smart contracts, pre-compiled smart contracts and code written in the assembly language of the Ethereum Virtual Machine.
Link: https://www.nature.com/articles/s41598-023-32701-6

April 28, 2023