Quantum computers are prone to errors in their physical qubits. Error-correction systems group those components into logical qubits designed to retain information more reliably, a technique researchers view as essential for useful quantum computation, New Scientist explains.
What changed on Helios
The latest experiment used a color code, while Quantinuum's September test used a different Helix code on the same trapped-ion hardware. The company says the color-code result had one-tenth the error rate of its previous record and one-hundredth the best published result from another hardware type. Those comparisons come from Quantinuum rather than an independent replication.
The company also reports that every tested increase in code distance, a measure related to how many physical qubits protect the encoded information, cut the remaining logical errors by a factor of 12. In its tests, the Helix code used 10 physical qubits per logical qubit, while the color code used as many as 37 and delivered the lower logical error rate.
A milestone, not a finished machine
Quantinuum says the result puts Helios within the accuracy range targeted for Apollo, its planned universal fault-tolerant computer expected by 2029. Robin Blume-Kohout, co-director of Sandia National Laboratories' Quantum Performance Lab, was not involved in the study and said in the company's release that a scaled-up version could reach useful reliability with fewer than 1 million physical qubits.
That projection is not a demonstration at full scale. Quantinuum's own cautionary statement says the experiment may not reproduce in a larger system, in logical computation or in commercial use, and that Apollo may miss its planned timeline.
