Fujitsu's Diamond Spin Quantum Computer Prototype: What the 1.55 K Press Release Actually Shows

On September 8, Fujitsu announced the completion of a "diamond spin quantum computer prototype." The phrase "quantum computer" conjures racks, cryostats, and running algorithms — but the press release reveals that what was actually demonstrated is Rabi oscillations on a chip.

The Technical Approach

The qubits are SnV (tin-vacancy) color centers: in the diamond lattice, two carbon atoms are replaced by one tin atom, and the hanging electron spin serves as the qubit. An aluminum oxide photonic integrated circuit — a layer of optical waveguides — delivers control and readout light directly to the color centers. The two most demanding engineering steps were heterogeneous bonding of diamond and aluminum oxide (two very different materials), and thinning the diamond from several hundred micrometers down to a few hundred nanometers. Control uses a joint compilation of optical, microwave, and radio-frequency signals, with scheduling handled by an FPGA. Partners are QuTech at Delft University of Technology (collaborating since October 2020) and the University of Tokyo for diamond processing.

The Temperature Rhetoric

The most eye-catching line in the promotional material is "minus 271.6 degrees Celsius" — sounding tantalizingly close to absolute zero (−273.13 °C). In Kelvin: 1.55 K. Compared to the 15–20 millikelvin operating points of superconducting qubits, that's roughly a hundred times warmer.

But it's worth clarifying what that buys. The real selling point of the diamond route was never "room-temperature operation" — the most common misreading in reprints — but that at 1.55 K, laser and microwave control suffices, without the massive layer-by-layer cooling apparatus of a dilution refrigerator. What's saved is an entire cooling system, a corner of the machine room, and a whole power bill; the thermometer reading itself is secondary.

As for the impressive numbers in the marketing materials — T2 above one second, two-qubit gate error rates below one in a thousand — these all come from older papers on the nitrogen-vacancy (NV) system, published in 2025. The star of those results is not this SnV chip. Old system data, new system chip — a material substitution lies between them, and the numbers cannot be directly extrapolated.

Escalation Through the Reprint Chain

Reading the Japanese original and the English coverage side by side shows how a sentence inflates as it gets republished:

• Fujitsu's Japanese original: qualified as the world's first *in the SnV on-chip integrated form* — a self-reported claim.
• The English press release: qualifiers weakened.
• Quantum Computing Report's coverage: written up as having *deployed cloud access*, when the original describes a demonstration in a test environment.
• Later reprints dropped the partners' cautious quotes, including the QuTech director's note that scalability remains "a long and arduous journey."

The phrase "world's first" deserves a second look: a footnote states this is Fujitsu's own assessment from an internal survey conducted in September this year. What has been demonstrated: Rabi oscillations, spin frequency calibration, dynamical decoupling, plus gate operations in a cloud test environment. The list of what was *not* disclosed is longer: qubit count, T1, this SnV chip's own measured fidelities, any algorithm benchmarks. In one sentence: the prototype of a quantum computer has been built, but the quantum computer's report card has not been issued.

The Endorsement Chain Is Real, Though

Harsh critique aside, the academic foundation of this technical route is solid. Component-level peer-reviewed results exist: universal gate set error rates below one in a thousand (PRApplied, 2025); remote CNOT gate teleportation across cryostats (Nature Communications, 2026) — the key puzzle piece for the modular route; and fault-tolerant operation of a single color center (Nature, 2022). All three come from the Taminiau and Hanson groups at QuTech, all in collaboration with Fujitsu. The components are credible; the system is unproven — a distinction more important than "world's first."

Why Only Fujitsu Is Betting on Diamond Among the Giants

Looking at Fujitsu's overall strategy reveals a hedged bet. Its superconducting line, in partnership with RIKEN, is scaling from 64 to 256 qubits, with a 1,000-qubit machine due this fiscal year. The diamond line targets a multi-module prototype in 2027. The two lines plan to converge in fiscal 2030 — the superconducting machine stacking over 10,000 physical qubits to yield 250 logical qubits, with diamond modules attached via optical interconnects — targeting 1,000 logical qubits in fiscal 2035.

Among global giants, superconducting (IBM, Google), trapped-ion (Quantinuum, IonQ), neutral-atom (QuEra), and silicon-spin (Intel, Diraq) routes all have heavy backers. Diamond spin — this "sixth route" — is being pursued by Fujitsu alone. The logic is understandable: photonic interconnect modularity is the most painful point of the superconducting route, and color centers naturally emit light — born with the shape of a photonic interface.

The QuTech director's cautious quote, deleted from English coverage, is actually the most honest summary of the whole story: scalability remains a long and arduous journey. Three watchpoints: when this SnV chip's own measured results appear in a paper; whether the 2027 multi-module prototype can demonstrate two-module entanglement — the real hard benchmark; and whether Fujitsu's 1,000-qubit superconducting machine ships on schedule this fiscal year, as the credit collateral for that 250-logical-qubit promise of 2030.

---

Sources: Fujitsu press release (Japanese original and English version, 2026-09-08) and 17-page technical brief; Quantum Computing Report coverage (2026-09-08); Bartling et al., PRApplied 2025 (arXiv 2403.10633); Iuliano et al., Nature Communications 17:4694 (2026, arXiv 2601.04848); Abobeih et al., Nature 2022; QuTech partnership background (since 2020-10).