Friday, October 11th, 2024

Enjoy a nice cup of freshly brewed quantum news ☕️ 

Today’s issue includes:

• The Qiskit Code Assistant is an AI-powered tool using large language models to help users write high-quality Qiskit code more efficiently.

• The honeycomb Floquet code outperforms the surface code in spin-optical quantum computing architectures.

• Incorporating soft analog readout information in the decoding process of a superconducting qubit using a surface code reduces logical error rates.

• Plus, interconnected QPUs, random circuit sampling for benchmarking, the first quantum attack on security, and more.

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

🤖 IBM Debuts AI-Powered Qiskit Code Assistant for Effortless Quantum Development

QUICK BYTE: IBM announced the availability of the Qiskit Code Assistant in private preview, an AI-powered tool leveraging large language models to help users write high-quality Qiskit code more efficiently.

DETAILS

• The Qiskit Code Assistant integrates IBM Watsonx large language models to assist users in generating and refining Qiskit quantum code based on natural language prompts or partial code, making the development process more efficient and accessible.

• Designed for IBM Quantum Premium Plan users, the assistant is integrated with popular environments like Visual Studio Code and JupyterLab, and offers autocomplete functionality to streamline quantum code creation.

• IBM developed the Qiskit HumanEval dataset to assess the performance of quantum code generation models, showing that the granite-8b-qiskit model outperformed other state-of-the-art large language models for quantum programming tasks, achieving 46.53% on Qiskit-specific tasks.

• IBM plans to open-source key components, including the Qiskit Granite model and Qiskit HumanEval dataset, to encourage collaboration and

  advancements in quantum code generation models within the broader quantum computing community.

🐝 Honeycomb Floquet Code Sets New Benchmark for Highest Photon Loss Threshold in Photonic Quantum Computing Without Multiplexing

QUICK BYTE: Quandela, Sorbonne Universite, and others demonstrated in a recent study that the honeycomb Floquet code outperforms the surface code in spin-optical quantum computing architectures, achieving the highest reported photon loss threshold for photonic quantum computing platforms without large-scale multiplexing.

DETAILS

• The research compares the performance of two quantum error-correcting codes—surface code and honeycomb Floquet code—within the spin-optical quantum computing architecture, using consistent noise models. The honeycomb Floquet code outperforms the surface code in this tailored architecture.

• The honeycomb Floquet code achieves a photon loss threshold of 6.4%, surpassing the 2.8% threshold of the surface code, making it the highest reported threshold for photonic platforms without relying on large-scale multiplexing.

• The study highlights the efficient use of native spin-optical quantum operations like Repeat-Until-Success (RUS) measurements, which contribute to the fault-tolerance of the honeycomb Floquet code. The code's periodic weight-2 measurements simplify error detection while dynamically generating stabilizers.

• This exemplifies the importance of co-designing error-correcting codes, syndrome extraction circuits, and native hardware operations to enhance fault tolerance in quantum computing.

🧠 QuTech and Riverlane Achieve 6.8% Reduction in Superconducting Qubit Error Rates

QUICK BYTE: QuTech and Riverlane found that incorporating soft analog readout information in the decoding process of a superconducting qubit using a surface code reduces logical error rates.

^DETAILS.

• The study explores the benefits of using soft (analog) readout information, rather than traditional binary readout, in the decoding of a distance-3 surface code for superconducting qubits, showing improved performance when correcting bit-flip errors in quantum error correction experiments.

• The experiment was conducted using a 17-qubit superconducting device with flux-tunable transmons. The team performed repeated Z-basis stabilizer measurements, utilizing a combination of minimum-weight perfect matching (MWPM) and recurrent neural network (NN) decoders, to evaluate logical qubit performance.

• By using soft information in the decoding process, the logical error rate was reduced by 6.8% with MWPM decoders and 5% with NN decoders, showcasing the significant potential of soft decoding methods in improving quantum error correction fidelity.

Nu Quantum is developing the infrastructure needed to scale quantum computing by interconnecting multiple quantum processing units through a quantum network. Their approach consists of three components: a qubit photon interface to transfer quantum information, a quantum networking unit to route photons and create entanglement, and distributed quantum correction to form logical qubits. Nu Quantum is progressing on its LYRA project, which aims to productize its quantum networking technology, and the company is now seeking Series A funding to continue its growth and advance toward commercial-scale quantum computing.

Recent research from Google Quantum AI explores random circuit sampling (RCS) as a method to assess quantum computer performance in the presence of noise, demonstrating an increase in circuit volume at the same fidelity compared to 2019 results. It reveals two key phase transitions in the performance of quantum processors as noise strength and qubit count change, highlighting the potential for noisy quantum computers to outperform classical supercomputers. The study also confirms the reliability of RCS as a benchmarking tool and rules out the possibility of spoofing algorithms in the weak noise regime.

Chinese researchers have used a D-Wave quantum computer to carry out what they claim is the first successful quantum attack on widely used SPN-structured encryption algorithms, posing a potential threat to sectors like banking and the military. Although they have not yet cracked specific passcodes, the study demonstrates progress in breaching encryption methods using quantum annealing, a process that rapidly solves complex problems by bypassing traditional computational barriers. The researchers acknowledge that environmental factors and underdeveloped hardware currently limit the full cryptographic potential of quantum computing.

Researchers from the University of Chicago have developed a method to directly bond single-crystal diamond membranes to various materials, including silicon and sapphire, without intermediary layers. This technique enables the creation of ultra-thin, high-quality diamond membranes with minimal contamination, supporting quantum technologies such as quantum sensing and nanophotonic integration. These membranes have strong spin coherence for nitrogen-vacancy centers and are compatible with complex quantum applications, including biosensing, making them a viable platform for future quantum and electronic advancements​.

The Open Quantum Institute (OQI) has introduced 10 quantum computing use cases aimed at addressing global challenges aligned with the United Nations Sustainable Development Goals (SDGs), including food security, healthcare, and environmental sustainability. These use cases leverage quantum machine learning and simulations to optimize processes such as plant genomics, food delivery, antibiotic discovery, and water leak detection. OQI emphasizes international collaboration to explore these quantum solutions, with a focus on accelerating advancements in science and societal impact through quantum technologies.

LISTEN

—friday ambience—

ENJOY

Some quantum mechanics — could dark energy, long thought to be constant, actually change over time? Recent findings from supernovae and baryon acoustic oscillations suggest a tantalizing possibility: dark energy might be decreasing, which would shake the foundations of our current cosmological model. If true, this could explain the mismatch between quantum mechanics and cosmological observations, forcing us to rethink the role of vacuum energy in the universe's expansion. As more data from upcoming missions like Euclid and DESI arrives, the next decade may unlock even deeper mysteries about the universe's fate.

WATCH