Vice President, Deputy Chief Legal Officer, Intellectual Property and Privacy

Quantinuum says Helios results strengthen its path to Apollo. by Drew Jolly | October 19, 2026 Quantinuum says a new error-correction result on its Helios quantum computer gives it greater confidence in a 2029 target for Apollo, its planned fault-tolerant system. The company reported 99.9995% logical memory fidelity, equivalent to roughly five failures per million cycles. "This derisks the industry," Quantinuum CEO Rajeeb Hazra told The Wall Street Journal. "We now know the recipe for both creating good physical qubits and error-correcting them in a scalable way using the least number of physical qubits to create a logical qubit." Logical Error Rate scaling with color code distance. The green line indicates Quantinuum Helios' projected logical error rate. The pink line indicates a superconducting architecture's projected logical error rate. (Quantinuum) In its color-code experiments, Quantinuum reported an average twelvefold reduction in logical memory errors as code distance increased. Higher code distances allow stronger error protection by using more physical qubits. Extrapolating from those results, the company estimates that its approach could support 100 logical qubits using approximately 30,000 physical qubits, compared with roughly 900,000 for the superconducting color-code implementation it analyzed. "The paper demonstrates real, serious performance gains from quantum error correction - a logical qubit nearly 100x better than its component physical qubits," said Robin Blume-Kohout, co-director of Sandia National Laboratories' Quantum Performance Lab, in Quantinuum's announcement. Blume-Kohout was not involved in Quantinuum's study. In his interview with the Journal, Hazra cited chemistry research as an example of how improved error correction could support commercial applications. The Quantinuum CEO said that the current level of fidelity could support algorithms for making predictions about simple molecules, with the potential for more complex algorithms as accuracy improves. "We are focused on commercialization, and all the use cases we see have different requirements for fault tolerance," he said. "That's why it is so fundamentally important versus just a scientific pursuit." The technical roadmap also carries financial importance. Quantinuum raised nearly $1.7 billion in its June initial public offering, but reported a $65.4 million loss on $8 million in revenue in its most recent quarter. Before Apollo, Quantinuum plans to introduce Sol, 2D-grid-based quantum computer. Lessons from Helios are intended to guide its development. Hazra told the Journal that combining accurate physical qubits with scalable error correction gives Quantinuum greater confidence in its schedule. "That is what underpins our roadmap confidence of getting to full fault tolerance by 2029," he said. Read Quantinuum's in-depth blog post for more details.

Quantinuum said its Helios computer reached a logical fidelity of 99.9995% in error correction, equal to five errors per million operations. The company calls it a world record, 10 times better than its own mark from September 2026 and 100 times better than the best published result on other hardware, from Google Quantum AI. The Broomfield, Colorado company used a "color code" within its error-correction architecture on the same Helios hardware as its earlier Helix code result. Each increase in code distance cut logical errors by a factor of 12. Quantinuum says the accuracy is now in the target range for Apollo, its planned fault-tolerant quantum computer due by 2029. Chief executive officer Rajeeb Hazra said the company has "crossed the threshold necessary for exponential improvements in accuracy.

Morgan Stanley picks Quantinuum as its top quantum computing stock. Quantinuum's Nasdaq-listed shares swung from an $86 high to under $50 after its June 2026 IPO, before Morgan Stanley's October 16 upgrade reset the conversation around its $15.6 billion valuation. Morgan Stanley just told its clients which quantum computing company it thinks actually wins. The bank handed Quantinuum an 88% upside target and skipped right past IonQ to do it. Joseph Moore, the Morgan Stanley analyst who covers the sector, upgraded Quantinuum to overweight on October 15 and set a $62 price target, well above where the stock closed the day before. That makes Quantinuum the bank's top pick for pure-play exposure to quantum computing. IonQ, by contrast, only gets an equal-weight rating and a $43 target, despite the firm's own asset management arm holding a 7% stake in that company. Moore's case rests on a simple scorecard. He graded the major quantum players on four measures: technical execution, business progress, manufacturability and growth, and Quantinuum came out on top of all four. That is a notable call to make so soon after the company's own rocky start as a public stock. Quantinuum listed on the Nasdaq Global Market on June 4, pricing its IPO at $60 a share and raising $1.68 billion, which valued the company at roughly $15.6 billion. Shares popped to $68 on debut and climbed as high as $86 within the first month. Then the air came out. By October, the stock had slid below $50. It briefly traded under its own IPO price: a round trip that left early believers wondering if the quantum hype cycle had peaked before the technology did. Morgan Stanley is betting that the selloff was about sentiment, not substance. The bank projects the quantum computing market could generate more than $100 billion a year by 2040. The real inflection point lands between 2029 and 2030, as companies move from research contracts to production systems. For Quantinuum specifically, Morgan Stanley forecasts sales jumping from around $270 million in 2028 to $2 billion by 2030. That's a sevenfold increase in two years, and it would require the company to actually deliver on hardware promises it has not yet proven at scale. IonQ has moved quantum computing closer to a measurable commercial market after passing $130 million in 2025 revenue and raising its 2026 outlook. The next investor test is whether backlog and commercial demand can turn into repeatable growth. - quantum computing companies with measurable revenue growth - how quantum computing startups monetize their technology Quantinuum traces its roots to 2021, when Honeywell merged its quantum computing division with Cambridge Quantum, a British quantum software developer. Honeywell still owns roughly 54% of the company. That hardware-plus-software structure, built inside an industrial conglomerate rather than bolted together through acquisitions, is central to Moore's manufacturability argument: Quantinuum controls more of its own stack than most competitors do. IonQ, by contrast, uses a trapped-ion approach too, and has spent heavily on acquisitions to build out its roadmap. Morgan Stanley's own asset management arm owns 18.4 million IonQ shares, about 7% of the company, which makes the bank's decision to rate Quantinuum higher an unusually direct statement. You don't hold a stake in one company and then publicly tell clients to buy its competitor's stock instead unless the research genuinely points that way. Quantinuum has also leaned on partnerships to build credibility heading into this upgrade. The company has worked with Nvidia, whose venture arm NVentures invested in Quantinuum's funding rounds. Investors in its pre-IPO raises included JPMorgan Chase, Mitsui and Amgen: a client list that reads more like a hardware supplier's than a speculative startup's. None of that erases the risk sitting under Moore's $62 target. Quantinuum lost money in its first earnings report as a public company in August, and an 88% upside call assumes the market rewards execution years before the $2 billion revenue figure actually shows up. Quantum computing stocks across the board, including D-Wave, Rigetti and Quantum Computing Inc., have traded on narrative as much as on contracts signed. Frankly, a single analyst call doesn't settle whether trapped-ion machines beat superconducting ones, or whether either beats whatever comes out of a national lab in five years. What it does settle is smaller but real: one of Wall Street's bigger banks looked at the same public information everyone else has and picked a side. For a sector that has mostly traded on press releases and conference demos, that is itself news. Join the discussion. No replies yet. Start the discussion. Julian Lim is an entrepreneur, technology writer, and a researcher. He started JL Data Analysis after graduating from NUS in Intelligent Systems. Julian writes about technology innovations and entrepreneurship on Business Times, Asia Pacific Magazine and occasionally contributes to Startup Fortune.