Tuesday, October 8th, 2024

Enjoy a nice cup of freshly brewed quantum news ☕️ 

Today’s issue includes:

• AlphaRouter is a reinforcement learning and Monte Carlo Tree Search based solution for optimizing quantum circuit routing.

• Real-time, low-latency quantum error correction uses an integrated FPGA decoder on a superconducting quantum processor.

• A method for fast, accurate, and local temperature control using qubits to regulate heat flow in quantum systems demonstrates potential for use in quantum computers, heat engines, and sensors.

• Plus, a record fundraising round for quantum software, Europe’s largest sampling-based photonic quantum computer, global laser detuning, and more.

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

⚡️ AWS Solution Provides 20% SWAP Reduction for Scalable Quantum Circuit Routing

QUICK BYTE: Scientists from AWS Quantum Technologies present AlphaRouter, a reinforcement learning and Monte Carlo Tree Search based solution for optimizing quantum circuit routing, demonstrating up to 20% reduction in SWAP gate overhead and improving scalability and generalization compared to existing methods.

DETAILS

• Quantum circuit routing is necessary in quantum computing due to limited qubit connectivity. This routing often requires inserting SWAP gates, which are costly and prone to errors. AlphaRouter addresses the NP-hard problem of minimizing these SWAP gates while transforming high-level quantum programs into executable quantum circuits.

• AlphaRouter integrates reinforcement learning with Monte Carlo Tree Search to find an optimal sequence of SWAP gate insertions. The RL agent is trained to reduce routing overhead by modeling the problem as a Markov Decision Process, while MCTS efficiently explores the space of possible actions. The network architecture used in the RL agent adds to the system's scalability by allowing AlphaRouter to handle large and complex circuits.

• According to the preprint, AlphaRouter achieved a 10-20% reduction in SWAP gates compared to the state-of-the-art methods (e.g., SABRE), even on previously unseen benchmarks. Additionally, it reduced the linear scaling coefficient of SWAP gates by 15%, making it adaptable to various quantum computer topologies. Tests showed faster runtime performance by eliminating MCTS during the inference phase.

• The reduced SWAP gate overhead directly impacts the efficiency of quantum computing, especially in applications such as quantum optimization and quantum chemistry. Plus, AlphaRouter's adaptability to different quantum computers and its ability to generalize across benchmarks speak to its potential as a scalable and globally optimized quantum circuit compilation tool.

📢 Real-Time Quantum Error Correction: Riverlane and Rigetti’s FPGA Decoder Reduces Latency on Superconducting Quantum Processor

QUICK BYTE: New research from Riverlane and Rigetti demonstrates real-time, low-latency quantum error correction using an integrated FPGA decoder on a superconducting quantum processor.

DETAILS.

• Quantum error correction remains a challenge in quantum computing. A recent preprint from Riverlane and Rigetti focuses on minimizing the latency and backlog issues in decoding quantum errors in real-time through a scalable, fast-feedback decoding system that prevents slowdowns in quantum computations.

• The researchers used an FPGA-based Collision Clustering decoder integrated into Rigetti's Ankaa-2 superconducting quantum processor control system. The system decodes error data from an 8-qubit experiment with decoding times below 1 µs per round, using a 9-measurement round protocol, achieving a full decoding response time of 9.6 µs including control and communication latencies.

• The experiment showed logical error suppression as the number of decoding rounds increased, with error rates dropping from 28.1% to 20.5% as decoding rounds increased from 5 to 25. The integrated FPGA decoder operated at a mean speed of 0.44 µs to 0.79 µs per decoding round, effectively avoiding the backlog problem on the superconducting device.

• The real-time decoding system may support future experiments requiring logical branching, such as lattice surgery and magic state teleportation, relevant for the development of universal quantum gate sets.

❄️ Qubit-Controlled Temperature: Nokia, RIKEN, and Aalto Present Nanosecond Precision Cooling for Quantum Systems

QUICK BYTE: Nokia Bell Labs, RIKEN, and University of Aalto scientists present a method for fast, accurate, and local temperature control using qubits to regulate heat flow in quantum systems, demonstrating potential for use in quantum computers, heat engines, and sensors by precisely controlling temperatures on nanosecond timescales.

DETAILS

• Controlling temperature at the nanoscale is highly relevant for quantum technologies such as quantum computing and sensing, where thermal fluctuations can degrade device performance through decoherence and dissipation.

• The authors propose using qubits as thermal intermediaries between a quantum system and its environment, where their energy splittings are modulated in time to control temperature. The model can cool or heat another quantum system dynamically by adjusting the qubits' temperatures, using only a few qubits to achieve this control at subkelvin temperatures.

• The system can cool a quantum device from 180 mK to 120 mK on a timescale of about 20 nanoseconds using qubits such as superconducting flux qubits or spin qubits. The performance demonstrates precise and localized control over temperature, with temperature modulation following specific protocols such as square-wave or cosine drives.

• This may provide a scalable and fast method for temperature control for quantum technologies, offering a practical solution to managing heat in solid-state systems and ensuring stable quantum operations over extended periods.

According to the newly released TQI Quarterly report, private investment in quantum technologies slowed in Q3 2024, with a 25% reduction in funding rounds and a 60% decrease in dollar amounts compared to Q2, although research advances continued to drive industry growth. Highlights include Riverlane's $75 million Series C round for quantum error correction technology and Quantum Circuits, Inc.'s over $60 million Series B round to commercialize quantum systems with integrated error detection. Other notable developments include NATO's investment in Aquark Technologies and the collaboration between Microsoft and Quantinuum, which set a record for the number of entangled logical qubits with high fidelity.

Q-CTRL has raised an additional $59M in its Series B funding round, bringing the total to $113M, setting a record for quantum software fundraising. Led by GP Bullhound, the round includes investments from Lockheed Martin Ventures and NTT Finance, reflecting the company’s strategic role in quantum computing and quantum sensing for defense and commercial applications. The funding will accelerate Q-CTRL's development of quantum control software, expand customer engagement, and support quantum workforce development initiatives globally.

China's Anhui Quantum Computing Engineering Research Center is expanding its production capacity for superconducting quantum computers, increasing the number of systems assembled simultaneously from five to eight. The facility, responsible for developing the 72-qubit Wukong chip, is also working on next-generation chips with more qubits and improved stability to advance quantum computing applications across industries such as logistics and pharmaceuticals.

Researchers at Paderborn University in Germany have successfully built Europe’s largest sampling-based photonic quantum computer, called the Paderborn Quantum Sampler. This system uses photons to perform quantum computations and features a programmable interferometer to reduce optical losses, a common issue in photonic systems. The quantum computer, developed in collaboration with private firms and coordinated by Q.ANT, is intended to be used for applications such as protein folding and molecular state calculations in pharmaceutical research, with future accessibility planned via the cloud.

Toyota Research Institute of North America and Xanadu have launched a project to use quantum computing for materials science simulations, specifically targeting the design and optimization of complex materials for quantum sensors and energy technologies. The collaboration focuses on using quantum algorithms and embedding theory to simulate spin defects in 2D materials, such as a negatively charged boron vacancy in hexagonal boron nitride, which are key for quantum sensing applications.

Kipu Quantum and QuEra Computing have expanded their collaboration to provide PlanQK users direct access to QuEra's Aquila 256-qubit neutral-atom quantum processors, streamlining quantum computing workflows for businesses. The PlanQK platform simplifies the development and deployment of quantum solutions through automated software tools, useful for industries such as logistics, pharmaceuticals, and finance to leverage pre-optimized quantum algorithms for solving complex problems without requiring deep technical expertise.

Researchers at ParityQC and the University of Innsbruck have developed a method to control quantum computers using a single global laser detuning, eliminating the need for precise individual laser controls. This technique simplifies hardware requirements and enables scalable encoding of combinatorial optimization problems, such as the maximum-weight independent set (MWIS), through the strategic placement of auxiliary atoms.

Eviden has launched PQC HSMaaS, a post-quantum cryptography Hardware Security Module as a Service, providing the EU sovereign data security with the highest level of certification from ANSSI. This cloud-independent solution, hosted in resilient data centers in France and managed by French teams, supports encryption algorithms designed to withstand quantum computing threats. PQC HSMaaS ensures data sovereignty, regulatory compliance with the NIS2 Directive, and is available through a subscription model, providing businesses with secure, future-proof encryption solutions.

LISTEN

On the most recent episode of the Superposition Guy’s podcast, host Yuval Boger, CMO of QuEra, sits down with Shmuel Bachinsky, CEO and Co-founder of Quantum Transistors. Shmuel discusses the company’s focus on using diamond-based solid-state spin qubits combined with silicon photonics and CMOS control planes, aiming to build a scalable quantum processor. He contrasts this approach with the challenges of silicon-based qubits, emphasizing higher operating temperatures and the potential for better two-qubit gate fidelity. Shmuel shares insights on scalability, funding, and future growth, and much more

ENJOY

In a recent Quanta Magazine article, Lenka Zdeborová speaks on being a teenager in the Czech Republic and finding inspiration in Isaac Asimov’s Foundation series, where mathematical predictions could map the future of civilizations. This fascination led her to statistical physics, a field that explains the behavior of large, unpredictable systems, and eventually to theoretical computer science. Now, as the head of the Statistical Physics of Computation Laboratory at EPFL, Zdeborová explores how phase transitions in physics can help model the behavior of algorithms, including neural networks. Her dream? To uncover the "thermodynamics of machine learning," much like the 18th-century understanding of steam engines revolutionized physics.

WATCH

In honor of the Nobel Prize in physics announced today, an introduction to physics informed neural networks, which integrate physical laws, like PDEs, into neural networks by modifying the loss function so solutions adhere to known physics: