Welcome to the Quantum Realm. 

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

🙌 📱 Android users rejoice. You officially have access to Quantum Odyssey — the epitome of gamification in education while teaching you quantum computing, accessible to all. Plus, a collab between quantum and space science, quantum demons, addressing quantum inclusivity, and more.

🗓️ UPCOMING

Sunday, June 30 | QTM-X Quantum Education Series 5 of 10 Quantum Hardware

Monday, July 1 | TQN Quantum Safe Transition Working Group

📰 NEWS QUICK BYTES

🕹️ Award-winning quantum game lauches on Google Play: It’s happening. Quantum Odyssey, the award-winning no-code, no math, educational game from Quarks Interactive has launched on the Google Play store today. This game is tailored for absolutely anyone who is eager to explore the depths of quantum computing and algorithm design. The mobile version features over 12 hours of gameplay, 250 quantum computing puzzles, and narrated learning modules. If you don’t have an Android device, the game is set to launch on Steam later this year. Wishlist it here. 

👩‍🚀 HBKU and NASA collaborate on quantum for space: Hamad Bin Khalifa University's College of Science and Engineering is partnering with NASA to explore the intersection of quantum networking and sensors in space-based science. Additionally, the collaboration includes efforts in training, education, and workforce development in emerging quantum technologies.

👿 Charging quantum batteries with quantum thermodynamics: A 19th-century thought experiment by James Clerk Maxwell, which imagined a demon breaking the laws of thermodynamics, has been demonstrated within a quantum computer to charge a quantum battery. Using 62 qubits, physicists at the Okinawa Institute of Science and Technology implemented a procedure to create a temperature difference between two groups of qubits, not only effectively building a quantum battery but demonstrating a modified law of quantum thermodynamics.

🤖 Multiverse Computing wins EU funding for a new class of LLMs: Multiverse Computing has been awarded 800,000 hours of supercomputer time as well as funding by the European Commission's AI-BOOST program to develop a large language model using quantum and quantum-inspired technology. Their new software, CompactifAI, reduces the size and cost of training LLMs while maintaining accuracy. This initiative would create a new class of LLMs that address the computational and energy demands in AI.

🌏️ Bridging the gap for global participation in quantum: While the US, China, and the EU have invested over $50 billion in quantum technologies, participation from both low and middle-income countries remains a challenge. At Optica’s Quantum 2.0 conference, scientists from these countries discussed hurdles and potential solutions. Emphasizing local impact, they highlighted the importance of retaining talent, transferring skills, and leveraging accessible resources like cloud-based quantum computing to encourage progress in their home countries. The need for tailored solutions and the focus on technologies that address local problems were underscored as essential strategies for inclusive growth in the quantum revolution.

⚡️ NTT Research Advances Photonic Quantum Computing with Coherent Ising Machine: NTT Research in Japan is developing a photonic quantum computer called the Coherent Ising Machine, which uses optical parametric oscillators to solve optimization problems mapped to an Ising model. Unlike traditional gate-based quantum computers, CIM leverages global coherence through physics-based processing which has advantages in speed and energy efficiency. However, challenges remain in terms of hardware flexibility and connectivity. Commercial availability is anticipated by the early 2030s.

EY and Oxford RTI encourage responsible quantum innovation: A whitepaper by EY and the University of Oxford’s Responsible Technology Institute emphasizes the importance of responsible innovation and proactive risk mitigation in the development of emerging technologies such as quantum computing. Key points include the necessity of addressing engineering challenges and countering exaggerated claims to build public trust. It also includes survey findings that highlight the need for accurate and responsible communication to avoid the pitfalls currently seen in adjacent tech spaces such as AI.

Pakistan's National Center for Quantum Computing is closing the quantum divide: Pakistan is on the road to advance in quantum technology with the establishment of its National Center for Quantum Computing, despite facing challenges such as economic constraints. With the global quantum market expected to reach $106 billion by 2040, Pakistan's initiative is necessary to close the technological and economic gap. Addressing cybersecurity vulnerabilities as well as leaning into international cooperation are essential steps. Overcoming these hurdles would secure Pakistan's position in quantum, as well as lead to industrial growth, job creation, and national security.

☕️ FRESHLY BREWED RESEARCH

Topological order from measurements and feed-forward on a trapped ion quantum computer: Mid-circuit measurements and feed-forward on Quantinuum’s H1 ion-trap quantum computer were used to implement deterministic non-unitary dynamics which led to high-fidelity and constant-depth preparation of long-range entangled quantum states. This method reduces resource requirements and allowed for exploring complex topological orders and non-Abelian anyons in quantum systems. Breakdown here.

Agnostic Phase Estimation: An entanglement-based phase estimation protocol achieves optimal parameter estimation without prior knowledge of the system's properties, which was a major limitation in quantum metrology. The experimental validation using a superconducting quantum processor showed an improvement in quantum Fisher information. Breakdown here.

Trade-off between Gradient Measurement Efficiency and Expressivity in Deep Quantum Neural Networks: This establishes a fundamental trade-off between gradient measurement efficiency and expressivity in deep quantum neural networks and proposes the stabilizer-logical product ansatz to optimize this balance. The SLPA reduces the quantum resources required for training while maintaining high accuracy. Breakdown here.

Scalable tomography of many-body quantum environments with low temporal entanglement: A hybrid quantum-classical algorithm for reconstructing the influence matrices of complex many-body quantum environments is presented. Quantum measurement data is collected from probe qubits interacting with the environment, followed by a machine learning algorithm to build a matrix-product state representation of the IM. This demonstrated a more scalable and efficient reconstruction of IMs, which can then be used to model quantum transport and non-Markovian dynamics in systems with moderate temporal entanglement.

Scaling Quantum Computations via Gate Virtualization: The quantum virtual machine is a system designed to execute large quantum circuits with high fidelity on noisy and small quantum processors by using gate virtualization. The QVM introduces a virtual circuit intermediate representation and a compiler to optimize quantum circuits by breaking them into smaller fragments and reducing circuit depths. This system is evaluated on IBM’s 7 and 27-qubit QPUs to show its ability to scale circuit sizes up to double the QPU capacity while still improving fidelity and reducing circuit depths.

Towards View-based Development of Quantum Software: A view-based quantum development approach using a single underlying model is proposed to address the interdisciplinary nature of quantum computing and the diverse views of its stakeholders. It introduces a quantum IDE to support various quantum views, such as mathematical descriptions, quantum circuits, and program code, while ensuring consistency and facilitating collaboration. This approach would overcome existing limitations in current quantum software development tools by enabling bi-directional editing and maintaining coherence (ha) across different views.

UNTIL TOMORROW.

BREAKDOWN

Topological order from measurements and feed-forward on a trapped ion quantum computer

🔍️ SIGNIFICANCE: 

• Preparing long-range entangled states is relevant to applications for quantum error correction, topologically ordered phases, and lattice gauge theories. Traditionally, preparing these states requires deep quantum circuits, which are not practical for near-term quantum devices due to limited coherence times. This study explores using mid-circuit measurements and feed-forward on a trapped ion quantum computer to implement non-unitary dynamics and achieve deterministic preparation of these states in constant depth. This effectively overcomes the limitations of unitary dynamics and also reduces the overall quantum resource requirements.

🧪 METHODOLOGY: 

• 20 qubits were encoded in two hyperfine states of 171Yb+ ions on Quantinuum’s H1 programmable ion-trap quantum computer.

• All ions are initialized in a state where the Z-type stabilizers are satisfied.

• The X-type stabilizers are measured on odd plaquettes and project the state into an eigenstate of these operators. Both ancilla-based and ancilla-free methods were used in order to show flexibility in the measurement process.

• Based on the measurement outcomes, conditional single-qubit Z gates are applied to correct any detected errors and a lookup-table decoder is used to handle error correction.

📊 OUTCOMES & OUTLOOK: 

• The toric code ground state was successfully prepared with high fidelity and a high-quality state.

• The measured topological entanglement entropy was consistent with Z2 topological order which confirms the quality of the prepared state.

• The creation and manipulation of non-Abelian anyons verified their non-trivial exchange statistics through braiding experiments.

• Overall, this means that the team was able to deterministically prepare long-range entangled states with constant depth circuits which is significant for both quantum error correction and the study of topologically ordered systems.

BREAKDOWN

Agnostic Phase Estimation

🔍️ SIGNIFICANCE: 

• In quantum metrology estimating a parameter (such as a rotation angle) without prior knowledge about the system's underlying properties presents a significant challenge. Traditional phase estimation techniques require detailed knowledge of the unitary in question which limits real-world applicability where that information is not always readily available. By leveraging entanglement, it is possible to achieve optimal phase estimation even in the absence of prior knowledge.

🧪 METHODOLOGY: 

• A two-qubit superconducting quantum processor is used to demonstrate phase estimation without prior knowledge of the rotation axis.

• The probe qubit is initially entangled with an ancilla qubit and undergoes an unknown rotation about an unknown axis.

• The entangled pair is then measured in an entangled basis so that more information can be extracted about the rotation angle than would be possible with a single-qubit sensor.

• This method relies on the mathematical equivalence between certain entanglement-manipulation experiments and closed timelike curves, which allow the researchers to effectively choose the probe's initial state after the rotation has occurred.

📊 OUTCOMES & OUTLOOK: 

• The entanglement-based protocol achieves a quantum Fisher information that is higher than any single-qubit sensor could achieve. Specifically, the QFI about the rotation angle can be boosted by 50% compared to entanglement-free strategies.

• The protocol is effective regardless of the unknown rotation axis, which resolves the limitation of previous methods that required prior knowledge of the rotation axis.

• Using a superconducting quantum processor, the protocol was experimentally validated and verified that the entanglement-based method consistently outperformed traditional approaches. The experimentally measured QFI approached the theoretical maximum, confirming the effectiveness of the entanglement-based strategy.

• This is specifically relevant for quantum metrology and sensing and opens up new possibilities for more accurate and practical quantum measurements.

BREAKDOWN

Trade-off between Gradient Measurement Efficiency and Expressivity in Deep Quantum Neural Networks

🔍️ SIGNIFICANCE: 

• One challenge in training quantum neural networks is the efficient estimation of gradients. In QNNs, the quantum state collapses upon measurement which makes gradient estimation difficult. This study reveals that there is a fundamental trade-off between gradient measurement efficiency and the expressivity of deep QNNs. This trade-off is key to designing QNNs that can be trained efficiently while maintaining sufficient expressivity to solve complex problems.

• Previous methods focused on specific QNN architectures or used ad hoc techniques to estimate gradients. This paper instead provides a general theoretical framework that applies to a wide class of deep QNNs. The introduction of the stabilizer-logical product ansatz is a novel approach that uses the symmetric structure of quantum circuits to balance gradient measurement efficiency and expressivity. This general ansatz reaches the upper limit of the trade-off inequality which is a significant improvement over conventional methods.

🧪 METHODOLOGY: 

• The research was conducted through a combination of theoretical analysis and numerical demonstration.

• A general trade-off between gradient measurement efficiency and expressivity was derived using the dynamical Lie algebra framework. Gradient measurement efficiency was defined in terms of the simultaneous measurability of gradient components and expressivity as the dimension of the DLA.

• The stabilizer-logical product ansatz was proposed as a general QNN ansatz that can reach the upper limit of the trade-off inequality by using the symmetric structure of the quantum circuit. The SLPA uses stabilizer groups and logical operators to construct QNNs with gradient measurement efficiency.

• Numerical simulations were conducted to verify theoretical findings. They considered the task of learning an unknown symmetric function and showed that the SLPA reduces quantum resources required for training while maintaining accuracy and trainability compared to conventional QNNs.

📊 OUTCOMES & OUTLOOK: 

• The study established a fundamental trade-off between gradient measurement efficiency and expressivity in deep QNNs. More expressive QNNs require higher measurement costs for gradient estimation.

• The SLPA reaches the upper limit of the trade-off inequality and allows for for the simultaneous measurement of gradient components with the fewest types of quantum measurements for a given expressivity. The SLPA also significantly reduces the quantum resources required for training while maintaining high accuracy and trainability.

• A theoretical foundation for designing efficient QNNs is provided. By understanding and using the trade-off between gradient measurement efficiency and expressivity, researchers can develop more practical and scalable QNNs.