The Daily Qubit

The largest partially fault-tolerant implementation of QAOA, a privacy-preserving quantum machine learning system, teleporting logical qubits, and more.

Thursday, September 19th, 2024

Enjoy a nice cup of freshly brewed quantum news ☕️ 

Today’s issue includes:

  • JPMorgan Chase and Quantinuum scientists demonstrate a partially fault-tolerant implementation of the quantum approximate optimization algorithm.

  • HSBC and Quantinuum implemented post-quantum cryptography and quantum random number generators to secure their gold tokenization platform.

  • Researchers propose a new paradigm integrating federated learning, fully homomorphic encryption, and quantum computing to create a privacy-preserving machine learning system.

  • Plus, FTQC algorithms for aerospace, teleporting logical qubits, blind quantum machine learning, and more.

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

QUICK BYTE: A recently published arXiv preprint from JPMorgan Chase and Quantinuum scientists demonstrates a partially fault-tolerant implementation of the quantum approximate optimization algorithm using the Iceberg error detection code on a trapped-ion quantum computer.

DETAILS: 

  • Scientists from JPMorgan Chase and Quantinuum used the Iceberg quantum error detection code on a trapped-ion quantum computer to demonstrate a partially fault-tolerant implementation of the QAOA, improving the performance for problems with up to 20 logical qubits.

  • The Iceberg code protects against errors by encoding k logical qubits into (k + 2) physical qubits, improving algorithmic performance on smaller problem sizes , but struggles with larger qubit numbers.

  • The research proposes and calibrates a performance model that predicts how the Iceberg code functions, extrapolating the model to future hardware with improved error rates to determine the conditions needed for QAOA to outperform classical algorithms.

  • This study represents the largest demonstration of a practical quantum algorithm protected by error detection to date and emphasizes the need for near-term quantum error detection.

QUICK BYTE: HSBC, in collaboration with Quantinuum, successfully implemented post-quantum cryptography and quantum random number generators to secure their gold tokenization platform against potential quantum computing threats.

DETAILS: 

  • HSBC and Quantinuum demonstrated the first quantum-secure application for tokenized physical gold by using post-quantum cryptography and quantum random number generators on a distributed ledger technology platform, protecting the security of financial transactions against future quantum threats while maintaining system performance.

  • The proof of concept used a post-quantum cryptography virtual private network to transfer gold tokens across two distributed ledgers, proving interoperability between different blockchain platforms while enhancing security without impacting operational efficiency​

  • This trial is the first successful application of quantum-secure technology for gold tokenization, providing an example of how financial institutions can maintain security despite quantum threats without overhauling existing systems.

QUICK BYTE: Researchers from SVKM’s Dwarkadas J. Sanghvi College of Engineering, the GSSS Institute of Engineering and Technology for Women, New York University Abu Dhabi, NASA Ames Research Center and others propose a new paradigm integrating federated learning, fully homomorphic encryption, and quantum computing to create a privacy-preserving machine learning system.

DETAILS: 

  • Researchers from multiple institutions combine federated learning, fully homomorphic encryption, and quantum computing to create a new privacy-preserving machine learning architecture that demonstrates improved security and computation efficiency with minimal trade-offs in accuracy.

  • The overall federated quantum neural network incorporates fully homomorphic encryption for encrypted model updates while using quantum computing for accelerated processing.

  • The methodology includes the use of quantum layers and parameterized quantum circuits to train local models, which are encrypted and securely aggregated by a central server, with minimal loss of accuracy despite encryption-induced overhead.

  • This approach demonstrates the potential for improved privacy in machine learning models while also addressing the computational challenges of using FHE and quantum computing, relevant for secure AI development, especially in sensitive domains such as healthcare and finance.

Alice & Bob and Thales have partnered with the intention to develop quantum algorithms for fault tolerant quantum computers to accelerate the simulation of aerospace equipment, including radars and telecommunications antennas. The project will evaluate how FTQCs can exponentially speed up electromagnetic simulations and optimize aerospace designs, while determining the necessary resources and timeline for industrial-scale quantum computing solutions. With support from Inria, the collaboration will involve algorithm development, testing on aerospace equipment, and benchmarking performance, backed by a 2.6 million euro budget under the France 2030 plan.

Quantinuum researchers achieved the first-ever teleportation of a logical qubit using fault-tolerant methods on their H2 trapped-ion quantum processor, a notable development relevant for scalable quantum computing. They used two teleportation techniques—transversal gates and lattice surgery—both with real-time quantum error correction for high fidelity in the transfer of quantum states. This research holds practical implications for the development of quantum communication networks and scalable quantum systems.

A study from JPMorgan Chase and MIT introduces a new blind quantum machine learning protocol using a quantum bipartite correlator, designed for distributed quantum computing applications. The protocol protects privacy by using communication-efficient techniques to estimate the inner product of data held by two parties, without revealing the sensitive data. This method also reduces the communication overhead compared to classical protocols, making it ideal for privacy-preserving tasks such as linear regression in quantum machine learning. This foundation for secure, resource-efficient distributed quantum machine learning could be instrumental in fields such as finance and healthcare, where data privacy is high priority.

QCentroid and QPerfect announced a partnership to integrate QPerfect’s MIMIQ 1.0 virtual quantum computer into QCentroid’s QuantumOps platform. This collaboration would provide the tools for enterprises to design, test, and deploy large-scale quantum algorithms, while mitigating barriers such as cost and complexity

Researchers at University College Dublin and the Indian Institute of Technology Dhanbad have demonstrated how quantum interference in nano-scale electronic circuits can create "split-electrons," potentially producing Majorana fermions, a theoretical particle essential for topological quantum computers. Their study shows that by forcing electrons close together in nanoelectronic circuits, quantum interference alters their behavior, making them act as if they are split in two, offering a possible route to realizing and manipulating topological qubits.

Pasqal and Université de Sherbrooke have partners to support quantum computing education and research by co-developing hands-on educational materials, training a quantum-ready workforce, and encouraging joint research using Pasqal’s quantum technologies. The collaboration includes internship opportunities for students, the creation of a Research Chair in Quantum Computing at UdeS, and a focus on developing quantum software solutions to bridge the gap between academia and industry.

IQM Quantum Machines and the University of Oxford presented two new methods, Fourier Ansatz Spectrum Tuning Derivative Removal by Adiabatic Gate (FAST DRAG) and Higher-Derivative (HD) DRAG, in order to reduce leakage errors in single-qubit gates for superconducting qubits. The study demonstrates that these methods outperform conventional approaches through notable reduction in leakage error and improved gate speed, which could enable faster and more accurate quantum gate operations without requiring iterative optimization.

LISTEN

In this episode the Tech.eu podcast, Oxford Ionics co-founders Dr. Chris Ballance and Dr. Tom Harty discuss their quantum computing innovations, predicting a "ChatGPT moment" for the technology within the next 18 months to two years. They highlight Oxford Ionics' mass-producible quantum chips, recent government contracts, and the competitive landscape for talent.

ENJOY

Quantum computing, while once a distant dream, optimism abounds across the industry, especially in light of recent developments. According to a recent McKinsey roundtable with quantum industry leaders and academics, the growth of quantum companies and their march toward fault-tolerant quantum computing reflect the sector's increasing maturity with over 65% of experts anticipate fault-tolerant quantum computers by 2030. The symbiotic relationship between quantum computing and AI is eagerly being explored to discover new capabilities, while challenges such as quantum-era cybersecurity and talent development are being addressed to build a sustainable, innovative ecosystem.


WATCH

Quantum Brilliance CEO Mark Luo discusses their innovative use of synthetic diamonds for miniaturized, energy-efficient quantum computing with significant commercial potential:

iceberg code — you won’t believe the bugs that lurk beneath the surface. 📸: midjourney