Diamond & Quantum Weekly | Sep 7–13, 2026

The reporting window runs from 00:00 Beijing time on September 7 to 23:59 on September 13, 2026.

Weekly signal

The most important development this week was not another increase in the number of MPCVD projects. It was the appearance of two more substantive signals: diamond began to enter the engineering procurement interface as a product with selectable specifications, while the quantum field produced a system-level prototype for SnV diamond spin computing.

There is also a useful correction to the industry’s obsession with size. A new 12-inch disclosure is not, by itself, evidence of an industry transition.

1. Fujitsu presents an SnV diamond spin-computing prototype

On September 8, Fujitsu announced a working prototype of a diamond spin quantum computer integrating tin-vacancy, or SnV, centers with a photonic integrated circuit. The technical chain includes tin-ion implantation in diamond, bonding of heterogeneous materials, thinning the diamond from hundreds of micrometers to hundreds of nanometers, integration of the SnV centers with alumina waveguides, and conversion of general-purpose quantum gates into optical, microwave and RF control sequences.

The prototype operates at approximately 1.55 K and has been connected to Fujitsu’s Hybrid Quantum Computing Platform.

The significance is not the number of qubits. It is that the SnV route is appearing as a system prototype combining material, photonic devices, control and a platform. One important comparison remains necessary: the existing Fujitsu–QuTech NV route has reported a two-qubit gate fidelity of 99.93(5)%, and the QuTech team has reported an unconditional remote CNOT gate between independent NV diamond nodes. Those validated control results have not yet been reproduced on this SnV prototype. Moving from NV to SnV should therefore not be described automatically as an overall capability upgrade.

2. RFHIC DiaFlux becomes a specified CVD diamond heat-spreader product

RFHIC formally listed its DiaFlux 150/180 products this week. The polycrystalline CVD diamond grades are specified at nominal thermal conductivities of 1,500 and 1,800 W/m·K at 25 °C. The listed thicknesses are 200, 350 and 500 μm, with a tolerance of ±20 μm. Round pieces range from 3 to 100 mm in diameter, while square pieces range from 3 × 3 to 70 × 70 mm. The listed surface roughness is Ra <20 nm.

This is more meaningful than simply announcing that a company can make a diamond heat spreader. The product is now described in engineering-selection language: grade, dimensions, thickness, tolerance, roughness and form are all visible.

But “Production/Contact Sales” still does not mean that significant orders exist. The next evidence to watch is customer introduction, actual shipment, repeat purchase and batch-to-batch consistency.

3. CVD diamond and MPCVD

Domestic announcements this week included an EPC tender involving Zhengshi, a reported production milestone from Hami Qingtan, and a sample-stage update from Tanyue Advanced Materials.

Jiangxi Liufang Diamond disclosed progress on a 12-inch silicon-based diamond film and production-line commissioning, using an HFCVD route. The 12-inch figure itself is not the industry’s first. Sifangda had previously disclosed related 12-inch capabilities. The genuinely new element this week is the new participant and its HFCVD route.

Thermal conductivity, film-thickness uniformity, stress and warpage, contamination, yield definitions and customer validation have not yet been closed. It is therefore premature to use this disclosure as evidence of manufacturing maturity.

The signal worth retaining is broader: more participants are entering the field, but the industry’s most important missing evidence remains effective output, consistency, performance data, independent customer validation and repeat purchase.

4. Quantum technology

Beyond the Fujitsu SnV prototype, two methodological developments are worth recording.

An arXiv preprint proposed wide-field spin-noise imaging based on NV magnetic quenching without microwaves. It replaces the microwave link normally used in ODMR with continuous LED excitation and a low-frequency modulated magnetic field, and demonstrates spatial spin-noise imaging on material samples.

A related preprint from a Princeton team proposed momentum-resolved quantum-noise spectroscopy using an ensemble of diamond quantum sensors.

Both developments belong in a quantum research evidence base rather than in the industrial-news category.

5. Conferences and industry activity

The CIOE exhibition in Shenzhen, held September 9–11, was one of the week’s more concentrated display settings for diamond thermal management. Companies showed heat spreaders, windows, diamond-metal composites and optoelectronic cooling solutions.

The exhibition supports the view that applications are expanding toward optical modules, lasers and advanced packaging. It does not, by itself, establish customer adoption or production scale.

6. Social-media observation

International social platforms amplified the “1,000 logical qubits” figure associated with the Fujitsu prototype. The current official fact is a working prototype and a modular architecture; a future scaling target should not be written as present capability.

Accounts such as DIASEMI and SEMIXICON continued to post frequently about inventory, products and applications. They are useful indicators of supply-side marketing activity, but they did not provide new evidence of capacity, customers or a technology step-change this week.

7. What to watch next

1. Fujitsu and QuTech follow-up papers: SnV fidelity, coherence time, photon-collection efficiency and modular-interconnect metrics.
2. RFHIC DiaFlux: first customers, actual shipments, thermal-conductivity test methods and batch consistency.
3. Domestic MPCVD projects: movement from equipment and planning announcements toward effective output, yield and independent customer validation.