TL;DR
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
Quantinuum reported 99.9995% logical fidelity on its Helios quantum computer using a color code, which the company says is a tenfold reduction in logical error rate from its previous result. The study reports that logical errors fell by a factor of 12 at each tested increase in code distance; independent assessment, test conditions and practical performance at larger scale remain key questions.
Quantinuum says its Helios quantum computer achieved 99.9995% logical fidelity using a color-code error-correction method, a result the company describes as a tenfold reduction in logical error rate from its previous record. The October 19 announcement matters because suppressing errors in logical qubits is a necessary step toward larger quantum computations, but the reported result is not by itself a demonstration of a useful, fault-tolerant machine.
The company’s press release equates the reported fidelity to five errors per million operations. It says the result was produced on Helios using a color code within Quantinuum’s error-correction architecture, which combines trapped-ion hardware, all-to-all qubit connectivity, tailored decoders and hardware-level error correction. The company calls the result a world record; that characterization and the comparisons with other systems are Quantinuum’s claims.
According to the report, logical errors decreased exponentially as the code distance increased. At each increase tested, the remaining logical errors fell by a factor of 12, the study says. Code distance describes a code’s protection against errors and is connected to the number of physical qubits required. The report does not provide enough detail in the supplied material to independently assess the experiment’s full methodology or reproduce its comparisons.
The result follows a September 2026 demonstration on the same Helios hardware, when Quantinuum used its Helix code to report record-setting logical memory and computation. The two codes make different resource tradeoffs: the company says Helix used 10 physical qubits per logical qubit, while the color code used up to 37 and achieved lower logical error rates. The comparison suggests no single code necessarily offers the best balance of accuracy and qubit overhead for every task.
Error Suppression and Qubit Costs
Quantum information is vulnerable to noise, and error correction aims to protect it by encoding a logical qubit across multiple physical qubits. A declining logical error rate as code distance grows is an important sign that adding protection can improve reliability rather than merely add overhead. Quantinuum’s reported factor-of-12 reductions at tested distance increases speak to that technical goal, though they do not establish performance at every scale.
The tradeoff between lower error rates and physical-qubit overhead has practical consequences. The color code’s reported use of up to 37 physical qubits per logical qubit may be worthwhile where accuracy is the priority; Helix’s lower stated overhead may suit other workloads. How those differences translate into runtime, total resources and application performance has not been established by the supplied announcement.
Quantinuum says the result places its accuracy within the target range for Apollo, its planned universal fault-tolerant computer, which it expects to make available by 2029. That is a company roadmap, not a confirmed delivery date or evidence that Apollo’s full technical requirements have been met. Independent researcher Robin Blume-Kohout said the study shows meaningful performance gains, while his estimate about a feasible scaled system is a projection rather than a demonstrated result.
As an affiliate, we earn on qualifying purchases.
Helios Tests Two Error-Correction Codes
Quantinuum’s announcement follows its September report on the Helix code, also tested on Helios. The company presents both studies as evidence that its architecture can test different codes on the same hardware and adjust the balance between logical accuracy and resource efficiency. Using the same system offers a basis for comparing the approaches, though the supplied material does not include a full independent head-to-head analysis.
The latest work uses a color code, a known approach to quantum error correction. Quantinuum describes its broader architecture as supporting adaptive syndrome extraction, which adjusts error-checking measurements, and chain-mapping, which connects codes suited to different computational tasks. These are elements of the company’s technical approach; the announcement does not quantify how much each feature contributed to the reported fidelity.
Quantinuum also compares its result with a Google Quantum AI result on superconducting hardware, claiming its logical error rate is 100 times lower. That comparison is attributed to the company and would need to be read alongside the studies’ respective definitions, protocols and operating conditions. The supplied source does not include those details.
““We have reduced error rates tenfold in rapid succession using a different code on the same commercial hardware, demonstrating the flexibility of our error-correction architecture.””
— Rajeeb Hazra, Quantinuum CEO
As an affiliate, we earn on qualifying purchases.
Evidence Still Needs Scaling
The supplied report summarizes Quantinuum’s announcement but does not provide the underlying paper’s full methods, data or experimental conditions. It is therefore not possible from this material alone to assess how the 99.9995% fidelity was calculated, the exact operation set covered, or how the result compares under matched conditions with other hardware platforms. Quantinuum’s “world record” and 100-fold comparison should be treated as company claims pending review of those details.
It is also unclear how reliably the measured error suppression will persist as systems grow, how much time and hardware the decoding and error-checking require, and how the codes perform on longer computations. The company’s stated Apollo target and the outside expert’s projection of a sub-million-physical-qubit system remain forward-looking; neither is a result demonstrated in this announcement.
As an affiliate, we earn on qualifying purchases.
Publication and Apollo Roadmap
Quantinuum points readers to its research paper on fault tolerance, which is the next source for evaluating the methods and data behind the announcement. Further results on larger code distances, longer computations and the resource demands of real-time decoding would help show whether the reported gains hold beyond the tested regime.
The company says it is combining its hardware, error-correction architecture and software platform toward Apollo, which it expects to make available by 2029. That schedule remains a corporate plan. Future demonstrations will need to show how the code tradeoffs affect complete workloads and whether logical error suppression can support sustained fault-tolerant computation.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did Quantinuum report on Helios?
Quantinuum reported 99.9995% logical fidelity using a color code on Helios, which it equates to five errors per million operations. The company says this is a tenfold reduction in logical error rate from its previous result.
What does the factor-of-12 reduction mean?
According to the study summary, each tested increase in code distance reduced the remaining logical errors by a factor of 12. The claim describes the tested increases; it should not be read as a guarantee that every future increase will produce the same reduction.
How did the color code compare with Helix?
Quantinuum says Helix used 10 physical qubits per logical qubit, while the color code used up to 37 and achieved lower logical error rates. That describes a tradeoff between qubit overhead and reported accuracy.
Does this mean fault-tolerant quantum computing is ready?
No. The result is a reported error-correction milestone, not a demonstration of a fully useful universal fault-tolerant computer. Scaling, performance on longer computations and total resource requirements remain open questions.
When does Quantinuum expect Apollo?
Quantinuum says its planned Apollo system is expected to be available by 2029. That is a company target, not a confirmed launch date.
Source: rss
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
