Monday, October 14th, 2024

Enjoy a nice cup of freshly brewed quantum news ☕️ 

Today’s issue includes:

• Hybrid quantum neural networks are investigated in the application to Earth Observation (EO) tasks.

• Chemical accuracy was proclaimed achieved on multiple NISQ quantum computers using a hardware-agnostic algorithm for chemical computations with 40-60 qubits.

• The ARQUIN framework simulates large-scale, multi-node quantum computers using a layer-based approach.

• Plus, another roadmap with 2030 as a key year, quantum kernel machines for LHC anomalies, an algorithm for qubit reuse, and more.

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

🌏️ ESA and Sapienza University Explore Quantum Integration for Improved Earth Observation Deep Learning Models

QUICK BYTE: Researchers from the European Space Agency and the Sapienza University of Rome investigate the impact of key design elements in hybrid quantum neural networks applied to Earth Observation (EO) tasks.

DETAILS

• Various quantum computing libraries, specifically Qiskit and PennyLane, are analyzed according to their performance in training HQNNs for Earth Observation, showing comparable accuracies but highlighting PennyLane's faster convergence due to its integration with PyTorch and GPU support.

• In evaluating the sensitivity of both classical and quantum-enhanced neural networks to different initialization values (random seed values), the team finds that HQNNs demonstrate higher accuracy and stability under certain conditions, especially in convolutional models, although quantum elements can introduce some sensitivity variability.

• The study also presents a new integration of quantum circuits into vision transformers (HQViTs) for EO. These show modest performance improvements over classical ViTs, implying that even basic quantum integration may improve classical deep learning models in complex remote sensing tasks.

• Overall, these insights into the practical application of hybrid quantum-classical models suggest that further development of quantum layers could improve the resilience and scalability of models in EO tasks.

🧪 Qunova Computing HiVQE Algorithm Demonstrates Chemical Accuracy Across Multiple NISQ Quantum Systems with 1,000x Efficiency Boost

QUICK BYTE: Qunova Computing announced they achieved chemical accuracy on multiple NISQ-era quantum computers using their hardware-agnostic HiVQE algorithm for chemical computations with 40-60 qubits.

DETAILS

• Qunova Computing announced the demonstration of chemical accuracy (below 1.6 millihartrees) across three different NISQ-era quantum computers, displaying the hardware-agnostic capabilities of their HiVQE algorithm, which was able to compute energy estimations for lithium sulfide and other molecules.

• During Quantum Korea 2024, Qunova successfully demonstrated its algorithm on a 20-qubit IQM machine for three consecutive days, producing consistent results, and achieved comparable accuracy on IBM’s Eagle processor and AQT’s ion-based quantum system.

• The HiVQE algorithm reduced computational resource requirements by over 1,000 times compared to traditional variational quantum eigensolvers, making it potentially scalable for real-world chemical applications using NISQ machines and as few as 40-60 qubits.

• According to the team, the key to their algorithm is in removing "Pauli word measurements" from the traditional VQE, simplifying the quantum computation process, and feeding the quantum results into classical computers for rapid, accurate energy calculations.

⛓️ ARQUIN Framework Simulates Large-Scale, Multi-Node Quantum Computers

QUICK BYTE: Researchers from 14 institutions, with the support of the Co-design Center for Quantum Advantage, the Department of Energy, and others, developed the ARQUIN framework to simulate large-scale, multi-node quantum computers using a layer-based approach.

^DETAILS.

• The ARQUIN framework simulates large-scale, distributed quantum computers by connecting qubits across multiple nodes, even between dilution refrigerators, focusing on superconducting quantum devices.

• Each institution contributed to a specific layer of the quantum computing framework, addressing challenges like microwave-to-optical transduction and distributed quantum algorithms, to simulate the interaction of quantum components on a system-wide scale.

• PNNL researchers created the simulation pipeline and Quantum Roofline Model, integrating all the research components to benchmark future multi-node quantum computers, a key step toward scalable quantum systems.

• Though no functional multi-node quantum computer has been built, this research provides a co-design roadmap for future quantum hardware and software, expanding into other projects such as HetArch, which further investigates different quantum architectures.

Quandela has released its 2024-2030 roadmap, with the intention to achieve fault-tolerant quantum computing by 2030, the first logical qubits expected by 2025, and a large-scale quantum computer assembly by 2028. The company’s strategy focuses on integrated photonics as it requires fewer components than alternative technologies and may scale more readily. Other notable key milestones include the launch of a second quantum computer factory by 2027, hybrid QPU-GPU solutions by 2025, and quantum computing libraries for developers by 2028. Quandela is also part of the the PROQCIMA program to help develop French-designed quantum computer prototypes by 2032.

The upcoming Quantum Code Challenge Hackathon by CTE Cagliari DLAB focuses on developing quantum algorithms for smart cities, inviting both beginners and experts in quantum computing to participate. Organized by #QItaly and CRS4, the event seeks to expand knowledge of quantum computing, encourage collaboration, and solve coding challenges that support the Cagliari Digital Lab project. Participants will work remotely from October 22nd to 25th, using the qBraid platform and mentorship guidance, with the final projects presented and judged on the last day. The event encourages individuals or teams to take on either beginner or advanced quantum challenges. Register by October 18th here.

A recent study shows that quantum machine learning may be used to detect anomalies in data from the Large Hadron Collider, potentially uncovering new physics beyond the Standard Model. Researchers from ETH Zurich, CERN, and IBM used quantum kernel machines and clustering algorithms on IBM’s quantum computers, demonstrating that quantum models can outperform classical methods in identifying rare events when more qubits and entanglement are used. Through unsupervised learning, the QML approach minimizes bias and increases the chances of discovering unexpected phenomena in high-energy particle collisions.

Researchers from Oak Ridge National Laboratory, SUNY, and others propose a quantum network architecture that mirrors the classical packet-switching design of the internet. It adapts traditional congestion control and active queue management protocols to quantum networks, using quantum datagrams to manage the transmission of entangled states between nodes. Simulations demonstrate the architecture’s ability to manage quantum memory decoherence and maintain end-to-end fidelity, indicating that classical networking tools may be effectively applied to quantum systems to build a scalable and reliable quantum network.

GidNET, an algorithm designed by scientists from the University of British Columbia, may be used to optimize qubit reuse in quantum circuits. By analyzing a quantum circuit’s directed acyclic graph and candidate matrix, GidNET efficiently identifies pathways for reusing qubits, reducing circuit width and runtime. According to a recent preprint, the algorithm outperforms existing methods such as QNET and Qiskit, achieving up to 21% circuit width reduction and up to 99.3% faster runtime for large circuits, potentially improving quantum computations on devices with limited qubit resources.

Researchers at Nanyang Technological University (NTU), Singapore, have discovered a method to produce entangled photon pairs using niobium oxide dichloride flakes just 1.2 micrometers thick, which is 80 times thinner than a human hair. This approach eliminates the need for bulky optical equipment by ensuring photon synchronization within the thin crystal flakes, relevant for integrating quantum technologies into compact devices.

LISTEN

On the most recent episode of The Quantum Spin by HKA, host Veronica Combs joins Celia Merzbacher, Executive Director of the Quantum Economic Development Consortium (QED-C). They discuss the evolution of the quantum industry, the importance of identifying gaps in technology, workforce, and policy, and how consortiums like QED-C are fostering collaboration to advance the field. They also touch on practical applications of quantum technologies, international expansion, and the upcoming reauthorization of the National Quantum Initiative Act.

ENJOY

In a recent interview with Verdict, Carmen Palacios-Berraquero, CEO of Nu Quantum, speaks on the essential infrastructure required for scaling quantum computers, highlighting the need for quantum networking to interconnect QPUs. She discusses Nu Quantum's unique approach with its qubit photon interface and quantum networking unit, and shares progress on their UK-backed LYRA project. Palacios-Berraquero is optimistic about the UK’s leadership in quantum innovation and foresees commercial-scale quantum computing by 2029, driven by advances in modular systems and quantum correction technologies.

WATCH

A panel featuring Jerry Chow, Sarah Sheldon, Michael Biercuk, Travis Humble, and Sabrina Maniscalco provided an overview of quantum computing advancements, focusing on algorithm development and near-term use cases: