QUOPS Scores Put Five Orders of Magnitude Between Today and Utility
A 10 September arXiv paper introduces QUOPS, measuring executable circuit size and throughput across Google, IBM, and Quantinuum processors and projecting a five order of magnitude gap before recognized utility scale challenge problems become reachable.
What changed
Researchers posted Benchmarking the computational power of quantum computers to arXiv on 10 September 2026 (arXiv:2609.12146), introducing QUOPS, the Quantum Universal Operations Performance System. QUOPS scores the largest random circuits a machine can execute above a fidelity threshold and the rate at which it runs them, enabling cross platform comparison on physical qubits.
Applied to leading 2026 processors, reported scores include Google Willow at 216 QUOPS and 2.0 × 10⁷ QUOPS per second, IBM ibm_boston at 204 QUOPS and 3.1 × 10⁵ QUOPS per second, Quantinuum H2-1 at 1,320 QUOPS and 353 QUOPS per second, and Quantinuum Helios-1 at 1,504 QUOPS and 303 QUOPS per second. Translating resource needs from recognized challenge problems, the authors estimate computational capability must grow by five orders of magnitude to reach utility scale targets, motivating continued fault tolerant investment.
Why it matters
Quantum roadmaps are diverging across superconducting, trapped ion, and logical qubit architectures. Without an architecture agnostic yardstick, capital allocators and R&D leaders compare marketing claims instead of executable circuit depth. QUOPS gives portfolio teams a shared vocabulary for when a machine is widening capability versus merely increasing qubit count.
The paper also benchmarks a simple fault tolerant logical processor on up to eight [[7,1,3]] encoded logical qubits using Quantinuum Helios-1, linking today's physical qubit scores to a projection path for fault tolerant generations.
Who is affected
Quantum program leads and CIOs funding hybrid classical quantum workflows should rebaseline vendor comparisons on executable circuit size at fixed fidelity, not qubit totals alone.
Algorithm researchers targeting chemistry, materials, or optimization need QUOPS translated into whether their circuits fit inside current capability cones.
Investors comparing IBM throughput leadership against Quantinuum circuit size scores must weigh rate versus depth tradeoffs explicitly.
What to do next
Map your target application's two qubit gate depth and required fidelity onto published QUOPS capability regions before signing multi year cloud commitments, and ask vendors for QUOPS style curves on the specific transpilation paths you plan to use.
What to watch
Whether IBM, Google, and Quantinuum adopt QUOPS in public roadmap updates, and whether independent groups publish challenge problem to QUOPS translations for drug discovery and materials workloads through year end 2026.
Sources
- Primary. arXiv, Benchmarking the computational power of quantum computers (10 September 2026, arXiv:2609.12146). QUOPS definition, cross platform scores, and utility gap projection.
- Secondary. Quantinuum, Introducing the Quantum Universal Operations Performance System: QUOPS (14 September 2026). Industry framing and capability region figures adapted from the paper.