Welcome to the Quantum Realm. 

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

🧮 What’s better than one-way quantum computing? Two-way quantum computing. It’s basic math, really. Plus, the largest physics conference in the world is on its way to California.  

🗓️ THIS WEEK

Sunday, June 23 | QTM-X Quantum Education Series 5 of 10: Quantum Hardware

📰 NEWS QUICK BYTES

🔗 The missing link: Zapata AI's new paper charts a path from today's noisy quantum devices to scalable fault-tolerant systems. The research highlights the potential of early FTQC algorithms to bridge the gap that exists between current and future quantum technologies.

🌴 All physicists clear your 2025 Spring Break: The American Physical Society is renaming the 2025 joint March and April meetings to the “APS Global Physics Summit,” aligning with the International Year of Quantum Science and Technology. Expect the largest physics conference ever, with over 14,000 attendees.

🖥️ Crypto-agility is the new agile: Organizations need crypto-agility to adapt to evolving cryptography standards and protocols, especially with quantum computing on the rise. IBM Quantum Safe's tools help track, manage, and transition cryptographic systems to stay secure.

🔬 Finland seeking partner, must have 300 qubits: VTT Technical Research Centre of Finland is on the hunt for a partner to develop a 300-qubit quantum computer, optimizing it for new material development. This state-funded project aims to bolster Finland's quantum tech leadership.

💌 To TGCC, From Pasqal with love: Pasqal delivered a 100+ qubit QPU to TGCC, pushing Europe forward in hybrid high-performance computing. Integrated with the Joliot-Curie supercomputer, this QPU will enable European researchers to innovate with hybrid quantum-classical applications.

☕️ FRESHLY BREWED RESEARCH

Exponential concentration in quantum kernel methods: A study in quantum kernel methods for QML shows that under certain conditions, quantum kernels can exponentially concentrate, which makes models trivial and independent of input data. The research emphasizes the need for careful design of data embeddings, global measurements, and management of noise to ensure efficient and meaningful evaluation of quantum kernels. Breakdown here.

Grover’s algorithm on two-way quantum computer: Two-way quantum computing is shown to improve the practicality and reliability of Grover's algorithm by using a CPT version of state preparation to achieve constant complexity and greater noise resilience. This method makes it particularly useful for solving NP problems in noisy environments. Breakdown here.

A quantum algorithm to simulate Lindblad master equations: A quantum algorithm for simulating Markovian master equations uses probabilistic applications of unitary channels and state preparation to reduce gate complexity and eliminate the need for ancillary qubits. This algorithm is particularly notable for its ability to handle time-dependent Lindblad equations and its applicability to different noise models.

Transversal CNOT gate with multi-cycle error correction: A transversal CNOT gate between two logical qubits is constructed using the repetition code with flag qubits on IBM devices. This work shows error suppression with increasing code size and multiple rounds of error detection.

UNTIL TOMORROW.

BREAKDOWN

Exponential concentration in quantum kernel methods

🔍️ SIGNIFICANCE: 

• Quantum kernel methods in quantum machine learning involve embedding classical data into quantum states and computing their inner-products. This method is thought to guarantee optimal model parameters due to the convexity of the training landscape.

• However, the study finds that under certain conditions, quantum kernel values can become exponentially concentrated which makes the model predictions on unseen inputs trivial and independent of the input data. This challenges the assumption that quantum kernels can always be efficiently evaluated.

🧪 METHODOLOGY: 

• Four sources of kernel concentration were identified: data embedding expressivity, global measurements, entanglement, and noise.

• Analytical bounds were derived for each source to demonstrate the conditions under which exponential concentration occurs.

• Numerical simulations were used to validate the theoretical findings, focusing on tasks such as binary classification and regression.

• The impact of noise was investigated by modeling hardware imperfections and assessing their effect on kernel concentration.

📊 OUTCOMES & OUTLOOK: 

• Highly expressive data embeddings lead to quantum kernels that are exponentially concentrated, making it difficult to distinguish between different input data points.

• Quantum kernels evaluated using global measurements also tend to concentrate exponentially, even with low-expressivity embeddings.

• For projected quantum kernels, high entanglement in the encoded states causes the kernel values to concentrate. This makes the model predictions unreliable.

• Noise in quantum hardware causes the quantum kernels to exponentially concentrate which is unfortunate news for current-state devices.

• Ultimately, certain features should be avoided in quantum kernel methods to ensure efficient evaluation and meaningful model performance. Specifically, problem-agnostic embeddings, high entanglement, and deep encoding schemes should be approached with caution. The study provides guidelines for designing embeddings that minimize these concentration issues in an effort to improve the generalization and reliability of quantum kernel-based models.

BREAKDOWN

Grover’s algorithm on two-way quantum computer

🔍️ SIGNIFICANCE: 

• Traditional quantum computing methods for Grover’s algorithm have limitations in controlling the final state of computations, which results in randomness and inefficiency. The 2WQC model addresses these issues by using a CPT version of state preparation resulting in a more precise control over the final state. This is useful for solving NP problems even in noisy environments.

🧪 METHODOLOGY: 

• In the 2WQC Grover algorithm, the oracle operation is modified to work in conjunction with an ancilla qubit. Instead of changing the phase of the desired states, an X gate is applied to the ancilla qubit controlled by the desired states, which leads to desired outcomes without multiple repetitions of the oracle and diffusion steps.

• The study includes a detailed analysis of the algorithm's performance under various noise models, such as bit flip, phase flip, phase damping, and depolarizing channels. The 2WQC Grover algorithm resilience to noise types is evaluated through simulations.

📊 OUTCOMES & OUTLOOK: 

• The 2WQC version of Grover’s algorithm has a constant complexity of O(1) in ideal conditions which is an improvement over the traditional O(√N) complexity. This means that the 2WQC algorithm can find the desired state in a constant number of steps, regardless of the size of the search space.

• The 2WQC Grover algorithm shows more resilience to various noise types. The algorithm shows higher resilience to bit flip noise and depolarizing channel noise compared to the traditional Grover's algorithm. The 2WQC Grover algorithm is completely resilient to phase flip and phase damping noise.

• These results show that the 2WQC Grover algorithm can be more reliably implemented in noisy environments as compared to the traditional approach.