Quantinuum reports Helios error-correction record, raising the bar for quantum computing

Quantinuum announced on October 19 that its Helios quantum computer achieved 99.9995% logical fidelity—five errors per million error-protected operations in the reported test. Quantum error correction uses multiple error-prone physical quantum bits to protect information; logical fidelity measures the reliability of those protected operations, not the success rate of every task the whole computer performs.^[1][2]

The company says the error rate is one-tenth of its September 2026 record and one-hundredth of the best published result across other hardware approaches, reported by Google Quantum AI on a superconducting system. Quantinuum says this supports its goal of fault-tolerant quantum computing by 2029—computing that can continue reliably while correcting errors—not that it has already reached that goal.^[1]

Longer, more complex quantum computations require errors to be kept from accumulating. If Helios can preserve this reliability as the system grows, users could complete more useful operations before errors spoil the result, strengthening the case for businesses to buy quantum-computing services. That is the route from this result to commercial value, but the reported metric alone does not demonstrate that ability to scale.^[1]

For competing providers, the result sets a higher reliability benchmark when seeking customers and research funding. Lower error rates do not automatically mean lower computing costs or more revenue: commercial value still depends on whole-task performance, the hardware resources required and pricing, rather than an error-correction record alone.^[1]

Impact

What to watch next

• 2029: Watch whether Quantinuum meets its fault-tolerant quantum-computing target. That would test whether this error-correction record translates into reliable full computations; the year is a company goal, not a guaranteed delivery date.^[1]