NEC Is Reported to Have Stopped Building Quantum Hardware. The Company Will Not Confirm It

by Margaret Ashworth · September 8, 2026

Senior Analyst, Frontier Tech Wire

NEC is reported to have stopped in-house development of gate-based superconducting quantum computing hardware. That report is unconfirmed by NEC, which declined to comment on it, and no outlet in the chain has published a company announcement. What exists is Japanese press reporting relayed by four English-language trade outlets over the past several days, and everything below should be read on that footing.

The sourcing deserves to be stated rather than laundered. Quantum Zeitgeist published on 5 September, citing Nikkei Asia. The Quantum Insider followed on 6 September, also crediting Nikkei Asia. Quantum Computing Report published on 7 September, saying the news was "first reported by Nikkei Asia." Tom's Hardware, publishing on 8 September, credits Diamond Online as the original outlet. Frontier Tech Wire has not read either underlying Japanese report, and the accounts do not agree on who broke the story.

Nor do they agree on when the halt happened. The Quantum Insider says the decision took effect at the "end of March," without stating a year; Frontier Tech Wire has not seen a document that fixes it to March 2026, and the widely used framing that this coincides with NEC's fiscal year end is an inference rather than something the reporting states. Quantum Zeitgeist, meanwhile, writes that NEC "revealed" the halt on 5 September — a characterisation the other accounts do not support, since Tom's Hardware has NEC declining to comment on the same question.

NEC itself has not confirmed an exit. Tom's Hardware reports that, responding to Diamond, "the company declined to comment on whether it had exited quantum computer development and said that it continues to evaluate practical applications and industrialization of quantum technologies as well as conduct proof-of-concept projects with customers." That is the outlet's account of NEC's position, not a statement NEC issued about hardware. It is compatible with having shut the hardware programme and equally compatible with not having done so, which is presumably why it was framed that way. Treat the halt as reported, not as announced.

If the reports are right, the symbolism is heavy. Tom's Hardware writes that "in 1999, NEC researchers demonstrated the world's first superconducting solid-state qubit, which became a foundation of superconducting quantum computing." Quantum Computing Report places the work at NEC's Tsukuba laboratories and calls the company "the first entity to demonstrate a controlled solid-state superconducting qubit." The Quantum Insider frames it as the first demonstration of the operation of a superconducting solid-state qubit. Frontier Tech Wire has not independently adjudicated the priority claim; it is reported consistently across all four accounts and attributed here to them.

The people moved before the programme did. Quantum Zeitgeist reports that Yasunobu Nakamura "demonstrated the first superconducting charge qubit at NEC in 1999" and "later became director of RIKEN's Center for Quantum Computing." Tom's Hardware reports that "a number of researchers involved with the program — reportedly including a key research leader — have since moved to Fujitsu, which continues to invest heavily in superconducting quantum computing hardware." Institutional knowledge in this field walks; it does not evaporate.

What is reported as stopped and what is reported as continuing are different things, and the distinction matters. Quantum Computing Report lists the discontinued work as "Physical Superconducting Hardware R&D" and "Gate-Based QPU Fabrication Projects." What it lists as continuing is "Simulated Annealing & Quantum-Inspired Algorithms" and "Commercial Combinatorial Optimization Services." The Quantum Insider similarly has quantum annealing and quantum-inspired systems running on conventional computers as the surviving lines.

That continuing list is not quantum computing in the sense the CHIPS funding announcements use the term. Quantum-inspired annealing runs on classical hardware; classical emulation is, definitionally, simulation. On the hardware side, Quantum Zeitgeist reports that in June 2023 NEC "began joint cloud-access research with Tohoku University on an 8-qubit annealing machine, with the explicit goal of hybrid operation alongside classical simulated annealing"; no larger NEC annealing device appears in any of these accounts, though their silence is not proof one does not exist. A company can keep selling optimisation services indefinitely without owning a qubit.

The stated reason, as relayed by the trade press, is commercial rather than scientific. Quantum Computing Report frames it as "extended commercialization timelines and capital investment requirements necessary to generate a near-term return on investment." No outlet reports a technical failure, and none reports a headcount, a budget, or a write-down.

Set that against the same week's news from Washington. On Tuesday the U.S. Commerce Department signed $100 million CHIPS agreements with Rigetti, whose machines are superconducting, and D-Wave, whose annealing systems are superconducting — the architecture NEC is reported to have set down on return-on-investment grounds. Two organisations looked at the same physics and reached opposite conclusions, because they are answering different questions. NEC, if the reports are accurate, was asking when a product would pay back a corporate R&D line. Commerce is not.

There is also a national-programme question the reports leave hanging. Quantum Zeitgeist writes that "NEC serves as project manager for a superconducting-circuit effort under Japan's Moonshot program, which aims to realize a fault-tolerant quantum computer by 2050" — stated in the present tense. Whether that role transfers, lapses, or continues on paper is not addressed in any of the accounts, and it is the single most consequential unknown here for anyone tracking Japanese quantum capacity.

Japan's public commitment, at least rhetorically, is not shrinking. Quantum Zeitgeist notes that "Japan announced plans to send 30,000 young scientists abroad to study AI and quantum research," and Quantum Computing Report points to IBM Quantum System Two deployments in Japan at Kobe and Kawasaki. Fujitsu continues to invest in superconducting hardware, per Tom's Hardware. The domestic ecosystem is not disappearing; one long-running corporate programme inside it is reported to have closed.

What nobody has published: how many people were affected, what NEC had been spending, whether any public money was involved in the discontinued work, and whether the intellectual property is being licensed, shelved or transferred. Until NEC says something on the record, those blanks stay blank.

The honest reading is narrow, and it starts with the caveat rather than ending on it: this is a report the company at the centre of it has pointedly declined to confirm. If it is accurate, a corporate research line that traces back to a 1999 demonstration has ended, and that tells you about the payback horizons large diversified companies will tolerate. It does not tell you that superconducting qubits do not work, and it is not a verdict on any listed quantum company's roadmap or a reason to buy or sell anything.