Quantum Computing Firms Advance Error Correction for Reliable Operations

On October 19, 2026, Quantinuum announced a world-record quantum error-correction result, achieving 99.9995% logical memory fidelity on its commercial Helios quantum computer[9][11]. This represents a tenfold reduction in logical memory errors compared with the company's previous record from September 2026[9][10].

These breakthroughs in quantum error correction are critical because they address the inherent fragility of quantum bits, or qubits, moving the industry closer to building reliable, large-scale quantum computers capable of complex computations[9][12]. Progress in reducing error rates and increasing logical qubit stability is essential for the development of fault-tolerant quantum machines, which could unlock new possibilities in fields like drug discovery and materials science[12].

The Breakthrough in Error Correction

Quantinuum's October 19, 2026 announcement highlighted that its Helios computer achieved a logical memory fidelity of 99.9995%, meaning only five uncorrectable errors occurred per million error-correction cycles[9][11]. This result is also more than 100 times lower than the best published comparison from other hardware modalities[9][10]. Robin Blume-Kohout, co-director of Sandia National Laboratories' Quantum Performance Lab, who was not involved in the study, stated that the paper demonstrates "real, serious performance gains from quantum error correction"[10][11]. Blume-Kohout added that a scaled-up version of this system could achieve the reliability needed for useful fault-tolerant computation at a feasible size[10][11]. Quantinuum's successive breakthroughs used different codes on the same Helios hardware, illustrating how its error-correction architecture and trapped-ion technology drive rapid accuracy gains[10][11].

Other Advancements in Quantum Reliability

Other companies also reported significant progress. On September 24, 2026, Infleqtion announced it had entangled 30 logical qubits, encoded in 80 atoms, on its Sqale neutral-atom quantum computer[14]. This achievement met the company's logical-qubit target for 2026[14]. The experiment involved grouping the 80 atoms into 10 blocks, with each block encoding three logical qubits[14].

Google Quantum AI published research in Science on October 15, 2026, demonstrating a technique called "magic state cultivation" on its Willow quantum processor[6][7][8]. This method reduced state-preparation errors by about 40-fold and yielded 99.99% logical fidelity[5][7]. Research scientist Emma Rosenfeld and quantum software engineer Craig Gidney noted in a Google Quantum AI blog post that this result addresses a problem where preparing these states could consume a large portion of a future machine's resources[5]. Google's demonstration moved a proposed way to reduce this burden from theory to hardware[5].

Earlier in the year, on February 27, 2026, discussions within the quantum community highlighted error correction as a foundational architecture for scalability, shifting focus from raw qubit counts to metrics like reduced error rates and improved logical qubit yield[15]. In August 2026, a research paper introduced "Cornucopia codes," a theoretical quantum error-correction approach for neutral-atom quantum computers, aiming to reduce physical qubit overhead[16]. IBM's 156-qubit superconducting processor, ibm_fez, was used in research where dynamic circuits allowed measurements and resets to be interspersed with quantum evolution[3]. Maika Takita, an IBM researcher, explained the challenge of repeatedly measuring a qubit while the rest of the processor remained operational[3]. IBM's Quantum Roadmap outlines a path to quantum systems using 200 logical qubits by 2029[4].

What we could not confirm

We could not confirm the specific error rates achieved by Infleqtion's 30 entangled logical qubits on its Sqale neutral-atom quantum computer[14]. The evidence does not specify whether these logical qubits are error-corrected to a particular fidelity.