EU Quantum Flagship Benchmarks Achieve Scalable Quantum Computing Performance Metrics

The pursuit of practical quantum computers demands standardised ways to measure progress, and a team led by Zoltán Zimborás from HUN-REN Wigner Research Centre for Physics, Attila Portik from Eötvös Loránd University, and David Aguirre from University of the Basque Country UPV/EHU now proposes a comprehensive set of key performance indicators to address this need. This work, developed within the EU Quantum Flagship, establishes benchmarks that move beyond assessing individual components and instead evaluate the holistic performance of quantum systems. The researchers introduce four core benchmarks, testing capabilities ranging from complex circuit execution to entanglement generation and the benefits of quantum error correction, and crucially, provide clear protocols and scalable evaluation methods. These indicators promise a transparent and fair framework for comparing different quantum hardware platforms and tracking the journey towards fault-tolerant quantum computation, representing a significant step forward in the field.

Surface Code Decoding For Superconducting Qubits

Researchers investigate the feasibility of constructing a quantum error correcting code capable of tolerating realistic noise, a significant challenge in building practical quantum computers. The team focuses on developing a code based on the surface code, a leading candidate for fault-tolerant quantum computation, but modifies it to improve performance under specific noise conditions. Their approach involves designing a decoder, an algorithm that corrects errors detected in the quantum system, optimised for the characteristics of noise found in superconducting qubits, the building blocks of many quantum computers. This decoder incorporates a refined weighting scheme, carefully calibrated to account for the correlated nature of errors in superconducting qubits, where errors on nearby qubits are more likely to occur together.

The team demonstrates that this improved weighting scheme significantly reduces the logical error rate, a measure of the overall error in the quantum computation, compared to standard decoding methods. They validate their findings through extensive numerical simulations, modelling the behaviour of a large-scale quantum system with realistic noise characteristics. The simulations demonstrate that the proposed decoder outperforms existing methods in correcting errors and maintaining the integrity of quantum information, paving the way for more reliable and scalable quantum computers. Furthermore, the study explores the decoder’s performance under different noise models, showing it remains robust and effective even in challenging environments, highlighting its potential for practical implementation.

Qubit Scaling, Error Correction and Benchmarking

This document provides a detailed overview of the current state of quantum computing, focusing on progress in hardware, error correction, and benchmarking. Significant progress is being made in increasing the number of qubits in quantum processors, with examples including Zuchongzhi 3.0 and advancements from IBM and Google. Research and development are happening across various qubit modalities, including superconducting qubits, trapped ions, silicon spin qubits, and photonic qubits. Quantum error correction is essential for building practical quantum computers, as qubits are inherently noisy and errors accumulate during computations.

The surface code is a leading candidate for QEC due to its relatively simple architecture and tolerance to errors, and creating logical qubits, error-corrected qubits, is a major milestone. Recent demonstrations show the creation of logical qubits with better-than-physical error rates, and efficient decoding algorithms like minimum-weight matching and sparse blossom are being optimized for speed and scalability. Shor’s algorithm remains a key benchmark for evaluating quantum computer performance, and there is a growing focus on application-specific benchmarks that reflect real-world problems. Metrics like quantum volume and circuit depth are used to assess overall performance, and techniques like direct fidelity estimation are being developed to reduce the overhead of benchmarking.

The field is also exploring methods for predicting quantum system properties from very few measurements. Key research areas include optimizing decoding algorithms, improving memory access patterns for the surface code, and mitigating signal crosstalk. Hybrid quantum-classical algorithms are a promising approach for solving complex problems, and automated compilation and optimization tools are essential for making quantum computers easier to use. Challenges remain in scaling qubit numbers while maintaining fidelity, improving coherence times, reducing error rates, and developing more efficient QEC codes.

Clifford Volume Benchmarks Assess Quantum System Performance

Scientists have developed a suite of benchmarks designed to rigorously assess the performance of quantum computers, establishing key performance indicators within the EU Quantum Flagship program. This work delivers a coherent framework for transparent and fair evaluation of quantum hardware, tracking progress towards fault-tolerant computation. The team introduced four core benchmarks, each accompanied by clearly specified protocols and scalable evaluation methods, to move beyond assessing isolated components and instead measure holistic system performance. One benchmark evaluates system performance through complex circuit execution, employing a Clifford Volume protocol as a scalable alternative to the Quantum Volume.

Experiments demonstrate that the CLV benchmark allows for characterisation of output states in a scalable manner, maintaining a volumetric aspect by considering both qubit number and computational depth. Researchers also benchmarked multi-partite entanglement by preparing and evaluating Greenberger, Horne, Zeilinger states, confirming that GHZ state creation and evaluation provides a volumetric benchmark readily extendable towards logical qubits. The team further investigated algorithmic primitives by adapting Shor’s period-finding algorithm, revealing a significant reduction in gate count by focusing on finding the period of maximum-cycle linear permutations. Finally, scientists quantified the benefit of quantum error correction, demonstrating that encoding quantum information across multiple physical qubits can reduce error rates, despite the resource overhead. The benchmark assesses the improvement achieved through error correction, providing a quantifiable metric for evaluating its effectiveness.

Quantum Benchmarks Assess Scalable Performance

This research presents a suite of benchmarks designed to rigorously assess the performance of quantum computers, addressing a critical need as these processors grow in size and complexity. The team developed four key performance indicators, each targeting a distinct aspect of quantum computing capability, and designed to be applicable to both current and future quantum technologies. These benchmarks move beyond evaluating isolated components, instead focusing on holistic system performance and scalability, crucial for tracking progress towards fault-tolerant quantum computation. The researchers accompanied each benchmark with detailed protocols, standardised evaluation routines, and a software development kit, ensuring reproducibility and fair comparison across different quantum hardware platforms.

Furthermore, the team made the protocols publicly available and collected sample data to encourage community-driven improvements to the benchmarking suite. The authors acknowledge that current benchmarks often lack the scalability needed to accurately assess larger quantum processors, and that a lack of standardised evaluation methods hinders meaningful comparisons. Future work will likely focus on refining these benchmarks as quantum technology advances, and expanding the suite to cover an even broader range of quantum computing capabilities. The researchers emphasize the importance of continued community involvement in improving and validating these benchmarks, ensuring they remain relevant and effective as the field progresses.