The Daily Qubit

Thursday, October 31st, 2024

Enjoy a nice cup of freshly brewed quantum news ☕️ 

Today’s issue includes:

• A Bayesian quantum neural network is designed to enhance power flow calculations in renewable energy-rich power grids.

• Scientists introduce quantum digital twins for use in hybrid quantum-classical computations to advance uncertainty quantification and distributed computing tasks on quantum systems.

• Plus, a quantum link across cities, real-time error correction, an advanced photonic quantum system, and more.

And even more research, news, & events within quantum.

🍃 Bayesian Quantum Neural Network Contributes to Renewable Grid Management

QUICK BYTE: A team of IEEE researchers developed a Bayesian quantum neural network, specifically designed to enhance power flow calculations in renewable energy-rich power grids.

DETAILS

• Power flow calculation is a model of complexity — variables such as voltage, current, and power flows across components of a grid must be quantified and monitored in order to prevent blackouts and other potential issues. While there are several commonly used classical methods, they can become computationally-prohibitive, especially in large-scale systems.

• Bayesian methods are probabilistic and adjust beliefs about parameters based on observed data. They are particularly effective in modeling uncertainty, such as that from renewable energy fluctuations. The team selects a Bayesian element as the end model needs to effectively generalize such that it continues to provide accurate results, even in the face of new data.

• The overall developed BQNN framework has three key components—Encoding, Ansatz, and Observation—incorporating data compression and entanglement-based calculations. The model’s effectiveness was evaluated using metrics like effective dimension (measuring model complexity) and generalization error bound (assessing predictive accuracy on unseen data).

• Numerically, BQNNs achieve lower normalized mean squared error (Vmse) across various IEEE test systems (6-bus, 30-bus, 118-bus), especially as renewable penetration increases. For instance, in the largest 118-bus system, the BQNN had a Vmse that was over 49% lower than traditional quantum neural networks under high renewable uncertainty scenarios.

• The outcome of the evaluation indicate that BQNNs are able to sufficiently capture complex power flow patterns comparable to, or in excess of, the performance of classical methods, while maintaining high accuracy even with unseen data. They may be further developed into a scalable solution for real-time power flow analysis in renewable-rich grids, showing promise for future applications in large-scale grid management.

👭 Digital Twins Used for Enhanced Uncertainty Quantification and Distributed Computing

QUICK BYTE: Scientists from the German Aerospace Center and ParTec AG introduce quantum digital twins for use in hybrid quantum-classical computations to advance uncertainty quantification and distributed computing tasks on quantum systems.

DETAILS

• Digital twins are virtual models that replicate the characteristics, behavior, and data of physical systems in real-time. A team considers the use of quantum digital twins to emulate real QPUs, capturing their noise characteristics to allow testing and optimization of hybrid quantum algorithms. The ideas is that the digital twins can provide an accessible way to evaluate quantum noise impacts on computations, especially useful when real QPUs are unavailable due to calibration or access limitations.

• Hybrid quantum ensembles—groups of hybrid quantum-classical models running in parallel on quantum digital twins—are used to assess prediction uncertainty. By combining quantum and classical layers, these twins effectively emulate distributed quantum systems.

• The team implemented their method on five quantum digital twins derived from IBM’s 127-qubit "ibm_sherbrooke" system, using metrics like quantum gate error rates and thermal relaxation time. The system demonstrated reliable predictions with uncertainty estimates by averaging results across multiple twin instances, validated using synthetic datasets for regression.

• Simulating parallel faulty QPUs not only aids in uncertainty quantification but also provides insights for refining quantum devices through noise analysis.

Rigetti and Riverlane demonstrated a real-time, low-latency quantum error correction system on Rigetti's 84-qubit Ankaa-2, achieving decoding speeds below the 1-microsecond threshold for quantum measurements. By integrating Riverlane’s fast-feedback control technology, it avoided computation backlogs. Rigetti's superconducting qubits, which reach gate speeds as short as 60-80 nanoseconds, make this integration possible.

A research team led by QuTech has successfully connected quantum processors across a 25 km distance between Delft and The Hague, using existing optical fiber infrastructure to establish a quantum link. This achievement involved a photon-efficient protocol and highly precise stabilization techniques to counter challenges like photon loss and ensure stable entanglement across long distances. Supported by European funding, this modular, scalable architecture lays essential groundwork for a European quantum internet, built around secure data exchange, distributed quantum computing, and frameworks for privacy.

Qolab, a quantum computing startup founded by former Google quantum leaders Alan Ho, John Martinis, and Robert McDermott, has secured $3.5 million from the Development Bank of Japan to advance its work on utility-scale superconducting quantum computers. With a focus on improving qubit coherence—essential for reliable quantum calculations—Qolab is partnering with Applied Materials to develop proprietary fabrication processes aimed at enhancing qubit stability and reducing error rates.

ORCA Computing has introduced the PT-2, an advanced photonic quantum system that integrates with high-performance computing infrastructures and NVIDIA’s CUDA-Q platform to enhance machine learning and generative AI workflows. The PT-2, deployed as a quantum testbed for the UK National Quantum Computing Centre, supports quantum-classical neural networks for applications across sectors like chemical formulation and vaccine development.

A study by researchers from NCSA and the University of Illinois has revealed low adoption rates of post-quantum cryptography, with only OpenSSH and Google Chrome currently implementing quantum-resistant encryption. Using a PQC Network Instrument paired with the Zeek network monitor, the team tracked adoption across real-world networks, finding significant obstacles such as complex protocol requirements and reliance on outdated cryptographic algorithms. The authors stress the need for broader PQC adoption, especially in protocols like RDP and DNS, to protect data against future quantum threats and mitigate vulnerabilities.

In the most recent episode of the superposition guy’s podcast Yuval Boger, Chief Commercial Officer at QuEra, is interviewed himself by guest host Jack Krupanski. Yuval shares his journey into quantum, discusses the recent Google investment, QuEra’s roadmap, the importance of government support, and the potential for a “quantum winter.” Jack and Yuval also discuss the evolution of quantum computing as part of the broader HPC ecosystem, expectations for production-scale deployment, challenges like talent acquisition, and much more.