Quantum Error Correction: New Codes Simplify Chip Design for Scalable Computers

A new set of tools for fault-tolerant quantum computing is being realised using Quantum Low-Density Parity-Check (qLDPC) “barbell” codes.

Shin Ho Choe of the IQM Quantum Computers and colleagues have designed these codes to overcome hardware limitations in implementing them on existing superconducting quantum hardware. The work details a scalable qLDPC code family with a chip layout supporting necessary two-qubit interactions without increasing hardware complexity as code distance grows. Simulations reveal these barbell codes can maintain information fidelity at a physical noise strength of 10-4 for trillions of Quantum Error Correction (QEC) cycles, utilising a modest overhead of fewer than 30 data qubits per logical qubit, and enabling fault-tolerant entangling gates between logical qubits. Sustained quantum error correction across trillions of cycles with a distance-14 barbell code For the first time, error rates dropped below 10-7, achieved using a distance-14 barbell code with a physical error rate exceeding 10-4. This enabled several trillion Quantum Error Correction (QEC) cycles, a feat previously impossible due to the rapid accumulation of errors inherent in quantum systems and representing a major leap in stability. Quantum information is notoriously fragile, susceptible to decoherence and gate errors, necessitating robust error correction strategies. Barbell codes, a new family of Quantum Low-Density Parity-Check (qLDPC) codes, accomplish this by encoding quantum information with a modest overhead of under 30 data qubits per logical qubit. This is a significant advantage, as lower qubit overhead translates directly to reduced hardware requirements and cost. A comparison with surface codes, a widely studied alternative, showed that, utilising 400 data qubits to encode 16 logical qubits at an error rate of 10-3, the barbell code exhibited a logical error rate per round of 8.8 times 10-7. This is nearly three orders of magnitude lower than the 9.6 times 10-4 achieved by a comparable surface code patch. Quantum information is notoriously fragile, susceptible to decoherence and gate errors, necessitating robust error correction strategies. The substantial reduction in logical error rate demonstrates the superior performance of the barbell code in preserving quantum information. Furthermore, the qubit overhead for these codes was up to eight times lower than that of surface codes with the same code distance, and the design requires only a hardware complexity metric of 1.65. This metric quantifies the number of connections needed between qubits, with lower values indicating simpler hardware requirements and easier implementation. The reduced complexity is crucial for scaling up quantum computers. Barbell qLDPC codes and long-lived quantum information preservation This advancement centres on a new chip layout designed to support qLDPC codes, a complex form of error correction similar to adding extra checks and redundancies to a digital file to ensure data integrity. In the quantum realm, this redundancy is achieved by encoding a single logical qubit, the unit of quantum information, using multiple physical qubits. The more physical qubits used, the more resilient the logical qubit becomes to errors.

The team engineered “barbell” codes, a specific type of qLDPC code constructed from pairs of X- and Z-tiles which form the basis of stabilizer measurements. Stabilizer measurements are crucial for detecting and correcting errors without directly measuring the encoded quantum information, which would destroy the superposition state. These tiles are translated across a lattice to create the code’s structure, defining the relationships between the physical qubits. Crucially, the design utilises near-local couplers, connections between qubits that aren’t directly adjacent, to enable syndrome extraction, a process where information about errors is gathered from the data qubits. Syndrome extraction allows the error correction process to identify the type and location of errors without collapsing the quantum state. Simulations assessed performance against circuit-level noise at a physical noise strength of 10-4, revealing similar logical performance per QEC round for both memory and logical multi-Pauli measurements. This provides insight into the code’s durability and resilience, demonstrating its ability to protect quantum information not only during storage (memory) but also during active computation (logical multi-Pauli measurements). The ability to maintain performance across both scenarios is vital for building a fully functional quantum computer. Barbell codes enable practical implementation of quantum error correction on limited-connectivity Stable quantum computation requires overcoming the inherent fragility of qubits, necessitating increasingly complex error correction schemes. Qubits are susceptible to environmental noise, leading to decoherence and errors in quantum operations. Quantum Low-Density Parity-Check (qLDPC) codes promise efficient encoding with minimal overhead, but a key hurdle remained: implementing these codes on real-world quantum chips with limited connectivity. Current quantum hardware typically allows only nearest-neighbour interactions between qubits, restricting the types of error correction codes that can be efficiently implemented. A solution is offered with “barbell” codes and a corresponding chip layout, although Dr. [Name] at [Institution] acknowledges a key limitation. Their simulations rely on a specific code family and noise model, which may not fully represent all quantum systems. Real-world quantum devices exhibit far more complex and varied errors, including crosstalk and variations in qubit properties. However, this establishes a key proof of principle. It demonstrates that qLDPC codes, considered a leading candidate for scalable quantum error correction, can be practically implemented on near-term hardware. This successfully demonstrates a scalable architecture for quantum error correction, utilising a new family of qLDPC codes and a corresponding chip layout. By carefully designing the connections between qubits, the researchers achieved constant hardware complexity regardless of how many qubits are used for encoding, addressing a key limitation of previous approaches. Maintaining stable quantum information for trillions of cycles, even with realistic noise levels of 10-4, signifies a substantial improvement in qubit fidelity and opens possibilities for more complex computations. This advancement is a crucial step towards realising fault-tolerant quantum computers, machines capable of performing computations beyond the reach of classical computers, with potential applications in drug discovery, materials science, and financial modelling. The ability to scale error correction while minimising hardware overhead is paramount for building practical and impactful quantum technologies. The researchers successfully demonstrated a scalable architecture for quantum error correction using a new family of qLDPC “barbell” codes and a corresponding chip layout. This is important because it addresses the challenge of implementing complex error correction on quantum chips with limited connectivity, maintaining constant hardware complexity as the code distance increases. Simulations showed these codes could preserve information for several trillion quantum error correction cycles with a physical noise strength of 10-4 using fewer than 30 data qubits per logical qubit. The authors suggest this establishes a key proof of principle for scalable quantum error correction on near-term hardware. 👉 More information 🗞 Barbell Codes: qLDPC Codes for Superconducting Quantum Hardware 🧠 ArXiv: https://arxiv.org/abs/2606.06062 Stay current. See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals. Tags: