The Daily Qubit

🎲 In a race to grab federal funding, it's Colorado vs Chicago's Bloch Tech Hub -- now taking bets.

Author

Cierra Choucair
May 14, 2024 • Estimated Reading Time: 12 minutes

Welcome to the Quantum Realm. 

Enjoy today’s breakdown of news, research, events & jobs within quantum.

I love to hear from you! Send me a message at [email protected] for musings, for fun, or for insight if it so appeals to you.

IN TODAY’S ISSUE:

  • RSA Conference 2024 has successfully closed and the Cryptographers Panel advises to keep calm, and carry on — but consider a hybrid cryptographic implementation, or secret key cryptography for long-term encrypted data

  • QAA combined with multistep probabilistic algorithms achieves quadratic acceleration for quantum state preparation

  • Hybrid quantum graph neural network approach for predicting the properties of complex materials

  • Plus, the D-Wave financial results are in, Chicago may have a worthy rival in Colorado for quantum federal funding, QuEra is recognized as a finalist in Fast Company awards, and SEALSQ Corp announces post-quantum semiconductor solutions for IoT

BRIEF BYTES

NEWS FOR THOSE IN A HURRY

TOP HEADLINES IN NEWS & RESEARCH

NEWS

Tags: PQC

BRIEF BYTE:  The RSA Conference 2024 addressed pressing topics like AI evolution, international policy impacts on technology, and quantum threats. The Cryptographers Panel at the 2024 RSA Conference addressed recent concerns about the security of lattice-based cryptography by urging calm, but implementation of secret key cryptography for long-term encrypted data.

WHAT: 

The 33rd annual RSA Conference in San Francisco hosted 41,000+ attendees, 650 speakers, and 600 exhibitors. Key discussions included the next evolution of AI, ethical considerations, security and privacy by design, and the implications of rapidly advancing technology on international policy. Highlights featured keynotes from industry leaders, panel discussions, and innovative startups. 

One panel discussion in particular addressed concerns around the April 2024 paper by Yilei Chen that claimed to use a quantum computer to find the shortest vectors in a lattice in polynomial time, potentially undermining lattice-based cryptography. Although an error was found in Chen's paper, the incident led to a reevaluation of the confidence in PQC algorithms under consideration by NIST.

Experts advised adopting a multilayer, hybrid cryptographic approach that combines PQC with current algorithms to strengthen encryption against future quantum attacks and current data harvesting. They also recommend avoiding public-key cryptography for long-term data protection and using secret key cryptography for data that needs to remain secure for over a decade.

RESEARCH

Tags: ALGORITHMS HYBRID

OVERVIEW OF QUADRATIC ACCELERATION OF MULTISTEP PROBABILISTIC ALGORITHMS FOR STATE PREPARATION

BRIEF BYTE: By combining quantum amplitude amplification with multistep probabilistic algorithms, researchers achieve quadratic acceleration for quantum state preparation.

WHY: 

  • The paper addresses a fundamental problem in quantum computing: the efficient and accurate preparation of quantum states, particularly ground states. Ground-state preparation is necessary for various quantum algorithms and simulations, but existing probabilistic methods do not offer significant computational speedups over classical approaches.

  • The traditional probabilistic algorithms use nonunitary operations with ancilla qubits to decay unwanted states probabilistically. However, these methods suffer from low success probabilities.

HOW: 

  • The authors propose an approach that combines quantum amplitude amplification with multistep probabilistic algorithms for quantum state preparation. By integrating QAA, the proposed method repeatedly amplifies the probability of obtaining the desired state which increases the overall efficiency of the probabilistic state preparation process.

  • A nonunitary operator is embedded in an extended unitary matrix, combined with forward and backward controlled real-time evolution operators. The application of QAA enhances the coefficients of the desired state through repeated operations, leading to a quadratic acceleration. The method is demonstrated using a PITE algorithm as an example.

RESULTS: 

  • The proposed method achieves quadratic acceleration in the computational process which demonstrates an improvement over traditional probabilistic algorithms and quantum phase estimation methods.

  • Numerical simulations using the one-dimensional Heisenberg model implemented on Qiskit demonstrate the strengths of the approach. The results show that the combined QAA and PITE method reduces the infidelity and computational cost compared to both standard PITE and QPE.

  • Specifically, the method exhibits exponential speedup in terms of infidelity reduction and quadratic acceleration in terms of success probability, making it exponentially faster than QPE for high-fidelity state preparation.

  • The method's ability to reduce the number of required ancilla qubits and the circuit depth makes it particularly suitable for early FTQC.

Source: Nishi, Hirofumi and Kosugi, Taichi and Nishiya, Yusuke and Matsushita, Yu-ichiro. Quadratic acceleration of multistep probabilistic algorithms for state preparation. Phys. Rev. Res. (2024). https://doi.org/10.1103/PhysRevResearch.6.L022041

PREPRINT

Tags: QML HYBRID

OVERVIEW OF HYBRID QUANTUM GRAPH NEURAL NETWORK FOR MOLECULAR PROPERTY PREDICTION

BRIEF BYTE: By integrating quantum feature encoding and quantum neural network layers with classical graph convolutional neural networks, researchers developed a hybrid quantum-classical model that improves upon the prediction of molecular properties, specifically formation energies of perovskite materials.

WHY: 

  • The research tackles the challenge of predicting molecular properties, specifically the formation energies of perovskite materials, using a novel hybrid quantum-classical graph neural network. Accurate prediction of these properties is critical in materials science for accelerating materials discovery and design, and traditional methods can be computationally intensive and time-consuming.

HOW: 

  • The researchers developed a gradient-free hybrid quantum-classical convoluted graph neural network (HyQCGNN) that combines classical graph convolutional neural networks with quantum feature encoding and quantum neural network layers.

  • They used amplitude encoding to represent the adjacency matrix of the molecular graph in a quantum state and implemented a trainable ansatz layer for quantum processing.

  • The optimization of this model was performed using non-gradient-based methods, specifically the Nevergrad optimization package.

RESULTS: 

  • The HyQCGNN model demonstrated competitive performance compared to traditional classical models like GENConv and XGBoost.

  • While the classical GNN achieved slightly better results in terms of R-squared values, the hybrid model showed promising results, indicating that quantum augmentation did not degrade the model's performance. This suggests that the integration of quantum layers can potentially enhance the predictive power of GNNs for complex materials properties.

Source: Michael Vitz and Hamed Mohammadbagherpoor and Samarth Sandeep and Andrew Vlasic and Richard Padbury and Anh Pham. Hybrid Quantum Graph Neural Network for Molecular Property Prediction. arXiv quant-ph. (2024). https://doi.org/10.48550/arXiv.2405.05205

EVENTS

JOBS POSTED WITHIN LAST 24 HOURS

UNTIL TOMORROW.

SUPPORT SCIENCE

Waking up before the world to dive into the quantum realm isn't just our job—it's our calling. And we're dreaming big with exclusive content for our community. If our work lights up your day, consider showing some love. Your support unlocks worlds—seen and unseen.

How many qubits was today's newsletter?

Login or Subscribe to participate in polls.

Interested in collaboration or promoting your company, product, job, or event to the quantum computing community? Reach out to us at [email protected]