Today’s Dispatches From Quantum:

Bose-Einstein…qubits? New molecular state has big implications for the future of quantum computing. Plus, the transformer model that’s transforming quantum state prediction.

Testing a new format to maximize focus on news and minimize distraction. Quick bytes on top, same long-form breakdowns below. Let me know what you think!

🗓️ THIS WEEK

Wednesday, June 5 | Quant Insights Conference: Quantum Computing in Quant Finance

Wednesday, June 5| Quantum Computing Workshop w/ Classiq

Thursday, June 6 | QaaS w/ Quantonix

Saturday, June 8 | Towards Practical Quantum Computing: Addressing Crosstalk and Circuit Optim w/ Washington DC Quantum Computing Meetup

📰 NEWS QUICK BYTES

⚛️ Longer calculations possible with molecular BEC: Physicists have created a Bose–Einstein condensate of molecules for the first time. Molecular BECs could form the basis of a new kind of quantum computer, where each molecule acts as a qubit. Breakdown here.

🔌 Circuits for large-scale superconducting computers: The National Institute of Advanced Industrial Science and Technology in Japan has developed a superconducting circuit capable of controlling numerous qubits at low temperatures via a single cable using microwave multiplexing.

📓 Goodby IBM Quantum Lab, Hello qBraid: qBraid has been endorsed by IBM as a recommended notebook environment following the closure of IBM Quantum Lab. qBraid offers pre-configured Python environments, seamless execution of Qiskit code, and support for multiple quantum computing frameworks, along with advanced features for coding, collaboration, and education.

🤝 The newest strategic alliance: Kipu Quantum and Quantum-South have formed a strategic partnership to bring quantum computing solutions to sectors such as telecommunications, banking, and logistics in Latin America and beyond.

⚙️ An interview with Kipu Quantum founder: Professor Enrique Solano, founder of Kipu Quantum, discusses the importance of creativity in overcoming hardware constraints, the distinction between quantum supremacy and quantum advantage, and the need for Europe to avoid technological dependency.

🎡 Watch this IQC semiconductor spin qubit presentation: Semiconductor spin qubits are a promising platform for scalable fault-tolerant quantum computers and quantum communications. This talk covers the process of converting quantum states from single photon polarization to single electron spin states in quantum dots, with an emphasis on finding conversion efficiency using photonic nanostructures.

🛡️ Finland-based IQM hosts US DoD: Heidi Shyu, the U.S. Department of Defense Under Secretary for Research and Engineering, visited IQM Quantum Computers in Finland to tour facilities and discuss potential collaborations with U.S. agencies.

☁️ New PennyLane tutorial for Covalent Cloud: This tutorial demonstrates running GPU-accelerated quantum circuit simulations on Covalent Cloud with a focus on quantum support vector machines.

🎨 Take a break with this fun scroll: WIRED presents an illustrative look at Microsoft’s Azure Quantum goals and an overview of quantum computing through interactive elements and beautiful graphics.

☕️ FRESHLY BREWED RESEARCH

RydbergGPT: A generative pretrained transformer called RydbergGPT was designed to predict measurement outcomes for a neutral atom array quantum computer. Using an encoder-decoder architecture, it inputs the interacting Hamiltonian and outputs qubit measurement probabilities. Breakdown here.

Quantum computation of frequency-domain molecular response properties using a three-qubit iToffoli gate: This study explores using the iToffoli gate for computing molecular response properties on near-term quantum hardware. The iToffoli gate significantly reduces circuit depth and execution time as compared to traditional single and two-qubit gates. Breakdown here.

UNTIL TOMORROW.

BREAKDOWN

Molecular BEC Realized For the First Time

🔍 WHAT HAPPENED: 

• Physicists have cooled molecules to create a Bose–Einstein condensate where hundreds of molecules lock into a single quantum state. BECs have been made with atoms since 1995 but creating one with stable molecules is a significant new achievement.

• Molecules interact in more complex ways than atoms, making them harder to cool to the required near-absolute zero temperatures. The research team used microwave fields to prevent molecule collisions and enable further cooling without significant losses. The result is a BEC of over 1,000 sodium–caesium molecules cooled to 6 billionths of a degree above absolute zero.

🚀 WHY IT MATTERS: 

• Molecular BECs could be used to create new types of quantum computers by using the molecules' quantum states as qubits. These qubits can potentially remain stable for longer periods which would allow us to complete more complex and prolonged calculations.

• The ability to control these molecular condensates can lead to a deeper understanding of quantum phenomena and material properties as well as the exploration of exotic phases of matter, such as supersolid phases and quantum droplets.

• This development could lead to the creation of materials that flow without resistance and other advanced quantum technologies. It opens new avenues for quantum simulation and advances our understanding of quantum states.

BREAKDOWN

RydbergGPT

WHY: 

• This study uses advanced machine learning techniques to address the problem of accurately predicting quantum states. Traditional quantum Monte Carlo methods are effective, but are computationally intensive and may not scale efficiently with increasing system sizes. The transformer model commonly used in natural language processing has the ability to generalize across different quantum states and offers a new approach that can potentially provide more scalable solutions.

HOW: 

• The research uses an encoder-decoder architecture tailored for Rydberg atom arrays. The encoder input represents the interacting Hamiltonian, while the decoder output corresponds to the measurements of the Rydberg occupation.

• The model is trained on datasets of varying lattice sizes, where the number of training tokens or measurements varies for each lattice size for a training set that covers a wide range of configurations.

• The training involves using physical estimators such as energy, x-magnetization, and staggered magnetization to benchmark the model's performance.

RESULTS: 

• The transformer model demonstrates good agreement with QMC estimates for energy and staggered magnetization in the ground state which indicates that the model can effectively capture the physical properties of the system at low temperatures.

• While the model trained on the smallest dataset struggles to generalize to larger system sizes, models trained on more extensive datasets show improved performance for larger grid sizes.

• The model had significant deviations from exact observables at high temperatures, especially for x-magnetization, since the model's architecture approximates probability distributions that are less accurate for well-mixed Gibbs states as opposed to pure states.

• The ability to generalize beyond the training data and accurately predict quantum states underscores the potential of transformer models (and AI techniques in general) to work alongside quantum computing applications.

BREAKDOWN

Quantum computation of frequency-domain molecular response properties using a three-qubit iToffoli gate

🔍 WHY: 

• Efficiently computing frequency-domain molecular response properties on current-state quantum hardware is difficult. Traditional methods using single and two-qubit gates result in circuits that are too deep for current quantum processors, leading to inaccuracies from decoherence and noise. This study shows how the use of a high-fidelity three-qubit iToffoli gate can reduce circuit depth and execution time significantly. Specifically, the iToffoli gate can have a ~50% reduction in circuit depth and ~40% reduction in execution time compared to traditional gate sets.

• This alternative method can lead to more accurate quantum simulations of molecular properties, which is useful for various applications in chemistry and materials science.

🧪 HOW: 

• A superconducting quantum processor was used to implement an LCU-based algorithm for computing transition amplitudes in diatomic molecules.

• HOMO-LUMO models of NaH and KH were used to benchmark the method. The active space consisted of the highest occupied and lowest unoccupied molecular orbitals.

• The circuits were constructed to calculate both diagonal and off-diagonal transition amplitudes. These circuits included the iToffoli gate and were optimized to reduce depth and execution time.

• Techniques such as randomized compiling and McWeeny purification were employed to mitigate errors during circuit execution and post-processing.

📊 RESULTS: 

• The iToffoli gate enabled a ~50% reduction in circuit depth and ~40% reduction in execution time compared to traditional CZ gate-based circuits.

• The experimental results showed that the spectral functions and density-density response functions obtained using the iToffoli gate had good agreement with theoretical values. Specifically, the spectral functions had a maximum peak height deviation of 10.6%.

• Combining randomized compiling with McWeeny purification led to substantial improvements in system-qubit state fidelities, achieving average fidelities of 96.0% in off-diagonal circuits.

• The study demonstrates the practical usage of a native multi-qubit gate in quantum simulations which would allow us to devise more efficient quantum algorithms in chemistry and materials science.