Quantinuum Error Correction Record Hits 99.9995% on Helios
Quantinuum's Helios hit 99.9995% logical memory fidelity, cutting errors about twelvefold per code-distance step and strengthening its 2029 Apollo roadmap.
Quantum computing companies publish roadmaps to show how they plan to move from today's noisy, experimental machines to systems that can solve useful problems. This hub tracks the milestones on those roadmaps and how well the hardware lives up to them.
The field has shifted from simply adding more physical qubits to making qubits reliable. The milestones that matter most now involve quantum error correction: encoding information across many physical qubits to create logical qubits that can tolerate faults. Researchers are also working on related techniques, such as preparing the special "magic" states needed for a full set of quantum operations, at lower cost and with fewer errors.
Several major players now point to similar windows later this decade for reaching fault-tolerant capability. That timeline affects more than labs. Security standards bodies and IT teams are preparing to move away from today's public-key cryptography, and the pace of quantum hardware progress shapes how urgent that work is.
Our coverage aims to separate real engineering progress from marketing, giving context on what each milestone shows and what problems remain unsolved.
Quantinuum's Helios hit 99.9995% logical memory fidelity, cutting errors about twelvefold per code-distance step and strengthening its 2029 Apollo roadmap.
Quantum X Labs began quantum error correction experiments on the Arbel processor at Israel's IQCC, run by Quantum Machines. Real-time QEC is still a future
NVision published a Nature paper on a single-molecule spin-photon qubit and partnered with HyperLight to put molecular qubits on lithium niobate photonic chips.
NIST published its final IR 8547 on Oct. 27, 2026. RSA and ECC are now deprecated at every key size after 2030, then disallowed after 2035.
NIST released its final post-quantum migration roadmap, retiring RSA and ECC by 2035, as Google reported a 40-fold error cut in quantum fault tolerance.
Quantum X Labs began error correction experiments on the Arbel processor at Israel's IQCC. Real-time correction and fault tolerance are later goals.
Quantum X Labs began error-correction experiments on IQCC's Arbel superconducting processor; it says real-time and fault-tolerant QEC are not yet demonstrated.
Quantinuum's 98-qubit Helios hit 99.9995% logical fidelity, about five errors per million cycles: 10x its September record and 100x Google's published result.
Google ran magic state cultivation on its Willow chip, cutting magic-state errors 40x at 99.99% fidelity and reviving debate over Bitcoin's Q-Day timeline.
Google cut magic-state errors 40-fold on Willow as IBM, Quantinuum and Microsoft aim for 2029 fault tolerance and US post-quantum crypto deadlines begin.
Google showed magic state cultivation working on its Willow chip, with 40x fewer errors and 99.99% fidelity. The result has restarted debate over Q-day
Google Quantum AI ran magic state cultivation on its Willow chip, cutting magic-state errors about 40x and reaching 99.99% T-state fidelity, per Science.
IBM and UChicago ran a 70-logical-qubit circuit on 97 physical qubits claiming verifiable advantage, but it relied on error detection and heavy post-selection.
IonQ says it has a real-time quantum error correction decoder running on one standard CPU, as new codes and roadmaps reduce the overhead of fault tolerance.
IonQ linked a trapped-ion qubit to a silicon-vacancy quantum memory at over 1,000 entanglements per second, beating the prior trapped-ion record by 4x.
A Jülich study of 10 quantum processors found Quantinuum's Helios-1 handled error-correction operations best, up to 91 data qubits, beating IBM and IQM.
In 2026 the quantum race moved to logical qubits: IBM ran 70, QuEra and Quantinuum posted disputed counts, and the big roadmaps all aim at 2029.
Infleqtion's 2025 neutral-atom redesign led to 30 logical qubits in 2026 as it targets 1,000 by 2030, but critics note the qubits rely on error detection.
Google cut magic-state errors 40x on Willow, IonQ unveiled a single-CPU decoder, IBM's Kookaburra is pending and Rigetti pushed back its 1,000-qubit goal.
IBM, Google and D-Wave push quantum error-correction milestones as the industry shifts focus from qubit counts to logical qubit reliability.
Reports of a Google 1,000-qubit processor lack support. Willow, with 105 qubits, is still Google's flagship, and error correction, not qubit count, is the real
Google Cloud says it will be fully ready for post-quantum cryptography by 2029. Harvest-now protections are due in 2027, and signatures and identity in 2028.
DARPA advanced IBM, IonQ, Atom Computing and Diraq to QBI Stage C amid new error-correction results, 2029 fault-tolerance roadmaps and 2030 PQC deadlines.
White House's August 2026 strategy cut the critical tech list to 14 areas, dropping data centers and batteries, adding post-quantum cryptography.
Google and Caltech spinout Oratomic slashed qubit estimates for breaking encryption in early April 2026 as error-correction roadmaps sped up.
Quantinuum and Infleqtion both raised their 2026 revenue outlooks in their first earnings as public companies, while their logical-qubit claims draw scrutiny.