The Million-Qubit Cabling Problem Is Real. FrostByte Just Got Paid to Try to Solve It.
The Cryo-CMOS Race to Control the Quantum Computer
Every superconducting quantum computer faces the same unglamorous problem: wiring. Each qubit needs its own control line running from room-temperature electronics down through the cryostat to the quantum chip. At tens of qubits, this is an inconvenience. At hundreds of thousands, it becomes a physical impossibility. The wires generate heat, consume space, introduce noise, and multiply cost. Two competing approaches are now racing to solve it.
The first is superconducting single flux quantum (SFQ) logic, which SEEQC demonstrated working at millikelvin temperatures in a five-qubit system, peer-reviewed and published in Nature Electronics earlier this year. The second is cryogenic CMOS — standard semiconductor transistor technology operated at cryogenic temperatures — and that is where FrostByte, a new Delft startup, is placing its bet.
FrostByte announced a €1.3 million seed round this week led by Graduate Ventures, with participation from InnovationQuarter, UNIIQ, Paeonia Group, and an undisclosed angel investor. The company is a spin-out from TU Delft and QuTech, widely considered among the world's leading centers for superconducting and silicon spin qubit research. The founders are James Kroll (CEO) and Luc Enthoven (CTO), with scientific advisors Fabio Sebastiano and Masoud Babaie — both internationally recognized pioneers in cryogenic integrated circuits.
That advisory lineage matters. Sebastiano and Babaie were among the authors of a landmark 2021 Nature paper (QuTech/TU Delft, using Intel's 22-nanometer FinFET process as a foundry) demonstrating a cryo-CMOS control chip operating at 3 kelvin. The chip generated tailored microwave bursts to drive silicon spin qubits, achieving single-qubit randomized benchmarking fidelity of 99.69% — statistically identical to room-temperature arbitrary waveform generators. That paper established cryo-CMOS as a viable path to integrated qubit control; FrostByte is its commercial descendant.
The technical problem FrostByte is attacking is the same one SEEQC addressed, but the engineering solution differs. Cryo-CMOS operates at 3 to 4 kelvin — warmer than the millikelvin temperatures at which superconducting qubits typically operate, but cold enough for standard CMOS transistors to function with dramatically reduced noise. The advantage is manufacturability: CMOS is a mature, dense, cheap semiconductor process with an existing global supply chain. The drawback is power dissipation at scale and compatibility questions with the qubit layer itself.
FrostByte's specific product focus, based on its public positioning, is cryogenic switches and ASICs designed to operate inside the dilution refrigerator at 4 kelvin and below — bringing the control architecture closer to the quantum processor. The "short-wire" approach, as it is known in the field, reduces the heat load and physical complexity that comes from routing room-temperature signals down long coaxial cables. The capital from this round will be used to expand the team and transition from lab prototypes to manufacturable hardware.
The competitive landscape is where this story gets interesting. Cryo-CMOS for quantum control is not a niche idea. SemiQon, a Finnish VTT spinout, raised €17.5 million last year to develop cryo-CMOS transistors capable of operating above 1 kelvin, claiming 0.1% of the power dissipation of room-temperature electronics at equivalent fidelity. SureCore, a UK-based company, has been building cryo-CMOS IP blocks and collaborating with foundries including GlobalFoundries, which also supports Quantum Motion's silicon spin qubit processors. Intel's Horse Ridge program has published on cryo-CMOS qubit control at 4 kelvin. IBM has demonstrated scalable cryo-CMOS systems for superconducting qubit control. Diraq and Emergence Quantum are pursuing the same path for silicon qubits, also with GlobalFoundries. The cryogenic CMOS controller market is projected to grow from $278 million in 2024 to $1.1 billion by 2033, according to one market estimate.
FrostByte's €1.3 million is pre-seed money — enough to build a small team and prove out prototype hardware, not to scale manufacturing. Graduate Ventures, which marked its 80th investment in five years with this backing, has a track record of investing early in TU Delft spinouts; it was also an early investor in QuantWare, the superconducting QPU manufacturer that went on to raise a €152 million Series B. That is the implicit promise in the FrostByte announcement: that this could become the next QuantWare-level story.
The honest caveat is that FrostByte has not published hardware. Its advisors' academic credentials are real, and the 2021 Nature paper is real. But a pre-seed company emerging from a research group is not the same as a demonstrated product. The question for FrostByte is whether the transition from prototype to productable chip can happen before the competition — and whether the "short-wire" approach survives contact with the thermal and electromagnetic realities of a working quantum processor at scale.
What FrostByte represents is the industrialization of a specific technical bet: that cryo-CMOS, not cryo-SFQ, will be the control architecture that lets quantum computers escape the wiring bottleneck. Two different temperatures, two different materials regimes, two different academic lineages. The market for cryo-CMOS controllers is real and growing. The startup that makes the transition from QuTech lab to reliable product will have solved one of quantum computing's least glamorous but most critical problems.