Independent study finds Quantinuum ahead of IBM on error-correction building blocks
Researchers at Germany's Jülich Supercomputing Centre tested the basic operations error correction needs on 10 machines from Quantinuum, IBM and IQM. Quantinuum's machines had the lowest error rates. The study isn't peer-reviewed yet.
A team at the Jülich Supercomputing Centre, which isn't affiliated with any of the companies, ran a set of tests on 10 commercial quantum processors from Quantinuum, IBM and IQM. The tests check the basic operations that error correction depends on: measuring qubits mid-calculation, resetting them, and acting on the results. They used circuit shapes taken from three different error-correction methods.
According to a summary of the paper by the trade publication Quantum Computing Report, Quantinuum's machines had the lowest effective error rates, several times lower than the best IBM processors tested. The researchers also found their score matched a real error-correction experiment on one IBM machine, a sign the test measures something meaningful.
How far the evidence has got: a preprint, not yet peer-reviewed. It's independent, which makes it valuable, but the detailed numbers should be read from the paper once it's reviewed.
Does this change the rankings? No. Rankings only change when QubitBenchmark publishes a checked result in its monthly data release.
Sources
- Evaluating the performance of QEC primitives on quantum processors at large width and depth, arXiv preprint 2610.05928 (Jülich Supercomputing Centre), 5 October 2026.