Quantinuum and Sandia Labs launch QUOPS, a new quantum benchmark

Quantinuum and Sandia National Laboratories have launched QUOPS, a new benchmark designed to move beyond qubit counts and assess what quantum systems can reliably do.

The Quantum Universal Operations Performance System measures computation size (“Q”) and operations per second (“Ω”) across diverse architectures, from noisy devices to future fault-tolerant machines. The developers aim to establish a common language for tracking progress toward practical quantum computing. QUOPS offers a standardized way to evaluate performance and inform procurement decisions as the field scales. QUOPS Measures System Performance Beyond Traditional Metrics The new Quantum Universal Operations Performance System, or QUOPS, establishes a system capability region by identifying the largest benchmark circuit size, designated “Q”, that a quantum computer can reliably execute, passing a pre-defined success threshold. This metric moves beyond simply counting qubits or measuring gate fidelity, instead focusing on demonstrable computational output, a shift important as quantum systems scale toward practical applications. QUOPS was developed by Sandia National Laboratories, with contributions from Quantinuum and NVIDIA, and is designed to be applicable across diverse quantum architectures and levels of fault tolerance, the company says. QUOPS measures not only the size of computation, Q, but also the speed at which it can be performed, reporting this as “Ω,” or effective operations per second. This two-dimensional assessment provides a more complete view of quantum system performance than single-point metrics, addressing a growing challenge for those evaluating quantum hardware. According to the developers, traditional component-level metrics remain essential for engineering, revealing control errors and connectivity constraints, but they are insufficient to determine what a machine can actually deliver in a complex computation. Quantinuum, a publicly listed company with about 630 people and $3.11 billion in funding, is actively pursuing fault-tolerant quantum computation with its trapped-ion technology, including a recently launched 98-qubit system named Helios. The architecture-agnostic nature of QUOPS is a deliberate design choice, intended to foster a common language within a fragmented field. The benchmark can be applied to systems employing different qubit modalities, error correction codes, and levels of fault tolerance, enabling more objective comparisons. This is particularly important as organizations consider deploying larger-scale quantum systems, requiring a clear understanding of computational capacity. The developers state, “As organizations move from experimentation toward larger-scale and potentially on-premise quantum systems, they need to know a simple thing: What computation can a machine actually execute successfully?” This highlights the need for a system-level measurement. QUOPS incorporates anti-gaming provisions, preventing vendors from optimizing solely for the benchmark rather than overall system performance. This feature is intended to provide buyers and governments with a more reliable basis for procurement decisions. The benchmark is designed to assess workloads requiring on the order of operations on thousands of qubits, a scale anticipated for future fault-tolerant quantum computers. Quantinuum’s recent partnership to explore quantum computing for energy challenges and its development of a backpropagation algorithm for quantum circuits demonstrate the company’s commitment to pushing the boundaries of quantum computation, and QUOPS provides a standardized framework for evaluating such advancements, according to the company. The system’s ability to reduce performance to a common currency, quantum operations, offers a clearer picture of capability than raw specifications alone. Q and Ω Metrics Define Computational Capability in QUOPS The newly introduced QUOPS benchmark quantifies quantum computer capability using two key metrics: “Q,” representing the largest benchmark circuit size a system can reliably execute, and “Ω,” measuring effective computational throughput. Unlike traditional evaluations focused on qubit count or gate fidelity, QUOPS assesses what a quantum computer can do, establishing a boundary of a system’s capability region through randomized workloads and success thresholds. QUOPS’ architecture-agnostic design allows for consistent evaluation across diverse quantum modalities, including Quantinuum’s trapped-ion technology and the superconducting systems from Google and IBM, the firm reports. Experimental data already demonstrates the trade-offs between these approaches; systems like Quantinuum’s Helios, with all-to-all qubit connectivity, achieve larger capability regions and higher Q scores but at a slower operational rate compared to the faster, though more limited, Willow and Boston superconducting processors. The benchmark specifies a random circuit construction, running circuits at varying widths and sizes to measure average fidelity against a predefined threshold, with each data point representing successful circuits and lines indicating capability limits. The significance of QUOPS extends to procurement and roadmap comparisons, offering a common yardstick for assessing quantum systems. Quantinuum, a public company listed on Nasdaq and backed by $3.11 billion in funding, has already measured its Helios system using QUOPS, alongside data from Google and IBM, as detailed in a recent scientific publication co-authored by the three organizations. QUOPS Enables Objective Quantum System Procurement The system-level approach considers the integrated performance of error correction, decoding, compilation, and other architectural factors, providing a more complete assessment than component-level measurements. This is particularly important given the current fragmentation of the field, where competing technologies make direct comparisons difficult. The development of QUOPS involved collaboration between Sandia National Laboratories, Quantinuum, and NVIDIA, highlighting the importance of shared standards in accelerating quantum progress, the company states. Quantinuum’s recent partnership with Synopsys exemplifies the growing demand for practical quantum solutions. QUOPS addresses a growing challenge for organizations considering larger-scale quantum deployments. As buyers and governments seek objective procurement decisions, they require a clear understanding of what computations a machine can actually deliver. The benchmark’s utility extends to roadmap comparisons, enabling HPC centers to assess when quantum computing will become beneficial for real-world workloads and allowing customers to define procurement thresholds based on demonstrable capability, such as running a trillion error-free operations. Saudi Aramco and Quantinuum began a partnership to explore quantum computing for energy challenges, a field where quantifiable performance metrics are essential for evaluating potential solutions. The benchmark’s design incorporates provisions to prevent vendors from optimizing solely for QUOPS scores, ensuring that improvements reflect genuine advancements in computational capability. This is important as organizations like Aramco explore quantum solutions for complex problems, requiring reliable and verifiable performance. QUOPS Benchmarks Quantinuum, Google, and IBM Processors The system also reports “Ω”, a metric describing effective throughput, providing a more complete picture of performance than single-value assessments. Quantinuum, Google, and IBM processors were among the first evaluated using QUOPS, revealing performance differences quantified by both Q and Ω values, and demonstrating the benchmark’s architecture-neutrality across trapped-ion, superconducting, and other modalities, by the company’s account. Data from the QUOPS scientific publication shows each labeled point representing experimental data from circuits passing a predefined threshold with high confidence, with lines indicating the filled capability limits of each machine, and stars marking the QUOPS score, the maximum circuit size successfully processed. A recent $100 million grant from the U.S. Department of Commerce, CHIPS Act, for trapped-ion manufacturing will further expand its capacity for research and development, and its R&D centre in Singapore, where Helios was deployed in March 2026, represents the first location outside the US to host the quantum computer. While QUOPS is not intended to replace all quantum benchmarks, it provides a valuable system-level reference point for comparison, flattening differences introduced by architectural choices and allowing for more informed procurement decisions. The benchmark’s design, as reported in the scientific publication, specifies a random circuit construction built for a specified width and size, with average fidelity measured and compared to a predefined threshold, offering a nuanced assessment of quantum capability. Source: https://www.quantinuum.com/blog/introducing-the-quantum-universal-operations-performance-system-quops More like thisQuantum Computing NewsSimulating material pairings could unlock topological quantum computersQuantum Computing NewsQuantinuum, Sandia Labs launch QUOPS to chart quantum computer powerQuantum Computing Business News$2 Billion US push aims to build quantum manufacturing baseQuantum Research NewsKimi Onoda Visits Delft to Deepen Quantum TiesStay currentSee today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals. Tags: Ivy Delaney Ivy Delaney has been working with neural networks and machine learning since the mid-nineties, back when a couple of hidden layers and a long afternoon of training counted as ambitious. She has watched the field go from academic curiosity to the thing quietly running underneath everything, and she brings that long view to quantum computing.

For Quantum Zeitgeist she covers the ground where the two fields meet. That means quantum machine learning and the variational algorithms it leans on, and it also means the less glamorous but more interesting story of classical machine learning already doing real work inside quantum machines, decoding error-correcting codes, calibrating noisy hardware and learning the error models that simulators depend on. She writes about the hardware those algorithms have to run on too, and about the post-quantum cryptography scramble that the same hardware has set off. Her stories typically start with the paper, whether that is peer-reviewed work, conference proceedings or an arXiv preprint, with the source linked so you can hold a claim up against the research it came from. She is unimpressed by benchmarks that will not say what they beat, and by demonstrations that only work in the press release.