Skip to content
hw.dev
hw.dev/signal/quantumdiamonds-91m-quantum-inspection-chips-2026
SignalQuantumDiamonds

QuantumDiamonds Raises 91M EUR to Scale Quantum-Sensing Chip Inspection Into Volume Production

QuantumDiamonds ships quantum NV-center inspection tools to US and Taiwan fabs, closing on a class of yield defects that optical and electron-beam tools cannot detect.

#testing#verification#semiconductor#manufacturing#ai-hardware
Read Original

QuantumDiamonds closed EUR 91 million (EUR 15M equity led by World Fund, EUR 76M from the EU Chips Act and German federal and Bavarian state funding) and is now in serial production with deployments live in US and Taiwan fabs. The constraint being removed is not inspection speed. It is the class of electrical defects that optical CD-SEM and electron-beam tools cannot see at all.

Diamond NV-center magnetometers map the magnetic field distribution inside a packaged or partially processed chip at room temperature, pinpointing current-carrying defect sites: voids, shorts, trapped charge. No destructive sample prep. No cryogenic cooling. The defects they find are real failures that will show up as early-life field returns, but they are invisible to the patterning-quality tools (ASML inspection, KLA, Applied Materials) that currently define the fab metrology stack. Yield loss from this defect population has been treated as a process tax with no direct observability; the only mitigation has been design margin, which costs area and power at every node.

The equipment market implication is structural. KLA and Applied Materials own patterning inspection. QuantumDiamonds is building a parallel column for electrical defect inspection, with a physical sensing mechanism those incumbents cannot replicate without building a quantum sensing division from scratch. As N2 and A16 learning curves drive up the per-wafer value at risk, the fraction of yield budget addressable only by quantum sensing grows. Chipmakers that begin deploying NV-center inspection in the next 12-18 months will have yield attribution data the rest of the industry is flying blind without.