Welcome to the Quantum Realm.

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

🎨 As quantum moves into data centers, engineers must flex their flair for creative solutions to overcome noisy data center environments and the need for cryogenic operating temperatures. Plus, D-Wave expands AI solutions, a quantum pipeline makes strides towards drug discovery, and an algorithm for automating the tuning of quantum dots across architectures.

🗓️UPCOMING

Tuesday, July 30 | TQN Quantum Programming Learning Sessions

Tuesday, July 30 | TQN Weekly Meetup - Quantum Synthetic Data for Machine Learning

📰QUANTUM QUICK BYTES

💻 The modern data center is both quantum and classical, but not without feats of engineering: A shift towards hybrid computing systems that integrate quantum and classical technologies has permeated the industry as the benefits of leveraging the strengths outweigh the reverse, in most cases. However, integrating these technologies is no easy feat as issues such as noisy data center environments, differing software architectures, and technical language barriers exist between quantum and classical engineers. Regardless, there is increasing demand for on-premises quantum computers to avoid cloud latency and build local quantum ecosystems. Efforts are underway to address physical and software challenges, including adapting quantum systems to data center conditions and creating efficient orchestration systems for hybrid workloads. The unique requirements of quantum computers, such as operating at cryogenic temperatures and frequent parameter re-tuning, will require creative approaches in manufacturing and system support. Ultimately, as quantum computing evolves, deeper integration with classical systems is expected.

🧬 Chinese researchers have developed a quantum pipeline with the potential for more accurate and efficient drug discovery: Chinese researchers from Tencent Quantum Lab, China Pharmaceutical University, and AceMapAI Biotechnology have created a quantum computing pipeline for drug discovery that uses simulations and calculations to address challenges in drug design, such as drug-target interactions and bond energy requirements. Validated through two case studies on anticancer drugs, beta-lapachone and sotorasib, the pipeline accurately predicted spontaneous reactions and strong covalent bonding. In addition to the more obvious benefits that come from efficient drug discovery, the pipeline could also expand access to advanced computing tools as drug design experts without a quantum background would also be able to use it. Further explorations and improvements will be needed to improve both accuracy and efficiency.

🔒 NTU Singapore launches the QUASAR program to advance quantum cybersecurity research: NTU Singapore is advancing research in quantum cybersecurity through the QUASAR program to develop effective cybersecurity technologies in response to the challenges posed by quantum advancements. In collaboration with the Technical University of Munich, the program focuses on quantum-safe systems, encryption techniques, and securing future Internet of Things and 5G devices. Research areas to be explored include quantum cryptanalysis, post-quantum encryption, and the integration of quantum security modules, with potential applications in critical infrastructures like telecommunications and finance. The program will be led by a top cybersecurity academic and will include funding for PhD students, post-doctoral researchers, and scholarships. The collaboration between NTU and TUM will also extend to other areas such as sustainability, health, space, AI, and additive manufacturing, with the potential to include a double degree program.

🔐 Businesses prepare for the inevitable change to security standards with NIST set to publish soon: The U.S. National Institute of Standards and Technology will soon release four finalized post-quantum cryptography algorithms which will set new standards for cybersecurity. Even though today’s quantum computers are not yet capable of breaking current cryptographic methods, it’s imperative that businesses proactively prepare for PQC implementation. Keyfactor's Chief Security Officer, Chris Hickman, advises that businesses should prioritize identifying and managing their cryptographic assets. Additionally, asymmetric cryptography and PKI will be the first areas impacted by PQC, so organizations should begin by locating certificates and managing keys. Globally, many regions will follow NIST's lead, with regional variations for data sovereignty and domestically developed algorithms. For IoT devices, constrained by hardware limitations, interim solutions and identity gateways will be necessary to bridge classical and post-quantum algorithms.

🔬 INRS-led projects receive $7.4 million from NSERC to advance quantum technologies: Funded with $7.4 million from NSERC's Alliance Advantage grants, INRS professors are pursuing three innovative projects in quantum communication, computing, and sensing through the integration of photonics. A project led by Professor Sharif Sadaf focuses on developing a semiconductor platform for on-chip quantum communication, while another project led by Professor Roberto Morandotti is around developing scalable and energy-efficient quantum communication, imaging, and sensing technologies, with applications in cybersecurity and biomedicine. The third project, also under Morandotti’s direction, is to create a scalable quantum photonic processor to demonstrate the potential of a quantum internet. These efforts are important for advancing information systems and quantum technologies in Canada while also encouraging the future workforce for these fields.

🌐 D-Wave is expanding its quantum AI solutions: D-Wave announced additions to its Leap quantum cloud service that will complement the integration of quantum optimization with AI and ML, with a specific focus on addressing AI/ML workloads such as pre-training optimization, efficient model training, and new AI business use cases. The initiative responds to increasing demand amidst the AI industry's escalating computing and energy costs by using annealing quantum computing to improve energy efficiency and optimization. Key development areas include Quantum Distributions for generative AI, Restricted Boltzmann Machine architectures, and integrating GPUs with quantum cloud services for AI model support. Notable customer use cases include improved biological data analysis by researchers in Julich, significant speed-ups in particle interaction simulations by TRIUMF, and fine-tuning RBMs by Honda Innovation Lab and Tohoku University.

☕️FRESHLY BREWED RESEARCH

CROSS-ARCHITECTURE TUNING OF SILICON AND SiGe-BASED QUANTUM DEVICES USING MACHINE LEARNING

QUICK BYTE: The Cross-Architecture Tuning Solution using AI algorithm automates the tuning of quantum dots in silicon FinFET, GeSi nanowire, and Ge/SiGe heterostructure devices and reduces tuning times from hours to minutes. CATSAI's adaptability across different device architectures and materials makes it suitable for scalable quantum computing as it’s compatible with existing semiconductor industry practices.

PRE-REQS:

• Tuning, as it relates to quantum computing, is the process of adjusting the operational parameters of quantum devices, such as quantum dots, to achieve optimal performance. This involves precisely setting gate voltages and other control variables to make sure that the qubits operate correctly, maintain coherence, and minimize errors, enabling reliable quantum computations.

SIGNIFICANCE: The use of silicon and silicon-based materials has been considered extensively for mass production of quantum computers as we already have established semiconductor industry practices to do so. Previous studies have shown that silicon and silicon-based devices can exhibit long coherence times and are inherently more scalable due our experience manufacturing silicon-based devices. Quantum dots, in particular, may be used in silicon for circuits with a large number of qubits.

However, quantum dots require precise control over their operational parameters. Tuning is a way to ensure that each device operates within the specific conditions needed for reliable quantum computing. In addition, tuning accounts for differences in manufacturing and environment and is necessary for large-scale integration. But tuning is time-consuming. This problem becomes more pronounced as we consider the combination of different device architectures. While previous algorithms have been developed to resolve parts of the tuning problem such as optimization and double quantum dots, none have explicitly taken into consideration how these techniques can be applied across different architectures or on scalable materials, such as silicon.

Researchers developed a machine-learning-based algorithm called Cross-Architecture Tuning Solution using AI that automates the tuning process and reduces the time required to do so from hours to just minutes. This was experimentally demonstrated by tuning quantum dots across three architectures: a 4-gate Si FinFET, a 5-gate GeSi nanowire, and a 7-gate Ge/SiGe heterostructure.

CATSAI differs from previous methods as it’s adaptable to different device architectures and material systems, and it incorporates advanced signal processing classification methods to handle charge switches and noise patterns. This versatility is important for developing quantum circuits at scale, as it ensures that diverse quantum devices can be tuned using the same algorithm. Overall, CATSAI tuned devices faster than human experts and random search methods as well as provided insights into the parameter space landscape for each device.

RESULTS: 

• The CATSAI algorithm reduced tuning times: achieved 30 minutes for a 4-gate Si FinFET, 10 minutes for a 5-gate GeSi nanowire, and 92 minutes for a 7-gate Ge/SiGe heterostructure, compared to hours of manual tuning

• The algorithm successfully tuned quantum dots across different device architectures and material systems as well as outperformed random search methods and human experts, providing more consistent and reliable tuning results across all tested devices

• CATSAI also provides insights into the parameter space landscape by characterizing regions where double quantum dot regimes are found; useful for informing future device design and optimization

HONORABLE RESEARCH MENTIONS:

Coreset selection is shown to improve the efficiency of quantum machine learning models. By selecting a smaller, representative subset of the training data, this technique speeds up the training process for quantum neural networks and quantum kernels without sacrificing accuracy. Models trained on these coresets perform as well as those trained on the full dataset, making QML more practical for large-scale applications. —> link to Coreset selection can accelerate quantum machine learning models with provable generalization

A hybrid quantum non-local neural network was used to improve image classification by capturing long-range dependencies using quantum entanglement. The QNL-Net outperformed traditional quantum classifiers and achieved near-perfect accuracy on binary classification tasks with the MNIST and CIFAR-10 datasets while using fewer qubits. —> link to A Scalable Quantum Non-local Neural Network for Image Classification

A method to increase qubit decoherence times uses stochastic resonance in two-level systems. Applying an oscillating field to a TLS shifts the noise spectrum to higher frequencies and increases qubit dephasing times. This technique, called noise source driving, is less studied than traditional dynamical decoupling and offers the advantage of manipulating the noise sources instead of the qubits themselves. The results indicate that NSD can be used to investigate the quantum coherence of TLS and optimize qubit performance by revealing the dephasing rates and energy level separations of the TLS. —> link to Using stochastic resonance of two-level systems to increase qubit decoherence times

UNTIL TOMORROW.