Error-correction record pushes quantum computers closer to usefulness — SkimNews
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- Quantinuum's Helios-1 quantum computer recorded the lowest error rate yet for logical qubits — as few as five errors per million qubit uses during storage testing, down from hundreds per million in past experiments.
- Helios-1 uses 98 ion-based physical qubits encoded via a "colour code" error-correction method, with improved laser pulse techniques correcting energy shifts that degrade computation accuracy.
- During a simple two-qubit mathematical operation, the error rate rose to about seven errors per 100,000 qubit uses — roughly 10x higher than the storage-only test.
- Error rates diminished twelve-fold when logical qubit size increased from 7 to 19 physical qubits, with similar gains from 19 to 37 — a critical signal that the method improves as systems scale.
- Steven Flammia at Virginia Tech called the result a course-setting step toward John Preskill's "megaquop" machine — 1 million reliable quantum operations — while Thomas Wong at Creighton University noted Helios-1's ~100 physical qubits fall short of the ~100 logical qubits needed for known useful applications.
- Quantinuum's next machine, Sol, will be nearly twice as large as Helios-1, while QuEra has announced plans to build a megaquop machine by 2028 using extremely cold atoms, and other teams already operate prototypes with more than 1,000 qubits.
Why it matters: Quantinuum now holds the lowest logical qubit error rate at 5 errors per million, but fault-tolerant computing still requires ~100 logical qubits and QuEra has announced plans for a megaquop machine by 2028. Quantinuum's Sol, nearly twice Helios-1's size, will be the next test of whether this error-correction lead holds at larger scale.
Ask SkimNews
