NEC, the Company That Created the Superconducting Qubit, Quits the Field After 27 Years

Forum topic · QianXun · 2026-09-08

Summary

NEC has quietly halted development of superconducting gate-based quantum computers, with hardware R&D ending around March 2026, according to a Nikkei Asia report on September 5, 2026. The decision carries historical weight: in April 1999, NEC researchers Yasunobu Nakamura, Yuri Pashkin, and Jaw-Shen Tsai published the Nature paper demonstrating coherent control of macroscopic quantum states in a Cooper-pair box, founding the superconducting qubit approach later pursued by IBM, Google, and Amazon. The foundational work was honored with the 2025 Nobel Prize in Physics. NEC is not leaving quantum technology entirely: it will keep commercializing quantum annealing, including the NEC Vector Annealing service running on SX-Aurora TSUBASA since August 2022 with 300,000 fully connected bits, and a 2023 superconducting parametron annealing project with Tohoku University, AIST, and ParityQC. The move follows similar pullbacks by Alibaba (2023) and Baidu (2024), while IBM, Google, Amazon, and Intel continue gate-based programs.

NEC, the company whose 1999 paper founded the field of superconducting qubits, has stopped developing superconducting quantum computers, Nikkei Asia reported late on September 5, 2026. According to sources, hardware development actually wrapped up at the end of March 2026. NEC issued no press release; Nikkei characterized the company as having apparently judged the return-on-investment horizon too long, and Jiji Press cited the view that profits were nowhere in sight and no hardware route had achieved dominance.

The 27-year arc

April 29, 1999 — Nakamura, Pashkin, and Tsai at NEC publish in *Nature* the first coherent manipulation of macroscopic quantum states in a Cooper-pair box, founding superconducting qubit research. The work underpins the 2025 Nobel Prize in Physics.
2003 — NEC and RIKEN demonstrate the first solid-state entanglement (*Nature* 425).
August 2022 — NEC launches Vector Annealing as a commercial service on its SX-Aurora TSUBASA vector supercomputer, with 300,000 fully connected bits.
2023 — Joint development of an 8-bit superconducting parametron annealing machine with Tohoku University, AIST, and ParityQC.
March 2026 — Superconducting quantum computer hardware development quietly ends.
September 5, 2026 — Nikkei Asia breaks the news; outlets worldwide follow.

Not a full retreat from quantum

What NEC stopped is the universal gate-based line only. The company is retreating to the segment that can generate revenue: quantum annealing. Annealing machines can take orders today, while gate-based machines still require error correction, scale, and a mature ecosystem. Translated, NEC's choice is simple: it stopped waiting.

A broader shakeout

NEC is not alone in stepping back from gate-based superconducting computing:

Alibaba closed its Damo Academy quantum lab in November 2023, donating equipment to Zhejiang University.
Baidu followed in January 2024, folding its team into the Beijing Academy of Quantum Information Sciences.
Intel remains active, deploying silicon spin qubit chips at Forschungszentrum Jülich in September 2025 and signing new agreements with Hitachi and NEDO in July 2026.
Google, IBM, and Amazon continue their gate-based programs.

The era when every major player aspired to a universal quantum computer is over; the list of contenders has simply gotten shorter.

The bitterest contrast

The entire significance of the 1999 paper was proving that a human-made circuit could exist in quantum superposition like an atom — no one asked when it would turn a profit. In the 2026 exit, every cited reason is financial. Science's journey from paper to product has a lower bound, and NEC's 27 years exceeded even its own patience.