A €5 million Finnish consortium is betting that the photonics powering the global internet can break the thermal barrier keeping quantum computers stuck at hundreds of qubits.
Quantum computers have to run at temperatures colder than deep space. The wiring that controls their qubits keeps the fridge warm.
That thermal contradiction is the reason every superconducting quantum computer on the planet still only works with tens to hundreds of qubits. Finland's new QScale consortium is betting €5 million that the fix is an unglamorous technology the world already runs on: the photonics that powers the global internet.
The project, announced by Tampere University on 1 June 2026, runs from 1 September 2026 through 31 August 2029 under Business Finland's "Rise to the Challenge" program. The VTT Technical Research Centre of Finland coordinates the work, with Tampere University and Aalto University as research partners. Tampere's share of the budget is roughly €1.6 million, and the project is built on the same telecom-photonics supply chain that has carried global communications for decades.
The problem QScale is trying to solve is rarely named in quantum press releases. Superconducting qubits live in dilution refrigerators a few thousandths of a degree above absolute zero. The control pulses that tell each qubit what to do have to be generated at room temperature and travel down electrical cables into the cold. Every one of those cables is a metal conductor carrying current into a system that is supposed to be the coldest place in the room. The cables themselves radiate heat at every stage, and each stage needs another layer of cooling to compensate. Multiply that by a million qubits and the cooling load becomes absurd.
Jukka Viheriälä, the QScale lead at Tampere University, framed the stakes bluntly. "If we can overcome this bottleneck, quantum computing can genuinely transition from the experimental stage to industrial-scale applications," he said in the project announcement. The QScale team's own back-of-envelope figure, drawn from extrapolating to a one-million-qubit machine, puts the cooling demand in the range of a nuclear power plant's output. That number is a scenario, not a measured benchmark, and QScale is a research consortium, not a power utility. But the order of magnitude captures why a thousand-qubit machine, let alone a million-qubit one, cannot be wired the way existing prototypes are.
The proposed replacement is to generate the qubit control signals optically, in the cold, rather than electrically at room temperature. Telecom-grade fiber carries data at light speed and at room temperature, and the microwave control pulses that drive the qubits can be synthesized from that optical signal right at the chip. The result is a control chain that adds almost no heat to the cryostat, scales with the same photonic components that already mass-produce the global internet, and lets engineers push the qubit count up without rebuilding the refrigerator.
Tampere's role in the project is the part closest to the physics. The team will work on high-frequency microwave signal generation for qubit control, on packaging that survives the cryogenic environment, and on the photonic-microwave interfaces that have to live next to the qubits themselves. The work will draw on Tampere's silicon photonics fab SiPFAB, the ChipIn and SoC Hub programs, and national nanofabrication and light-source facilities. Aalto contributes the device and quantum systems side. VTT holds the consortium together and runs the broader SUPREME project that QScale sits inside, which is meant to keep Finland at the front of the European quantum pack.
The Finland setting is not incidental. The Helsinki quantum cluster ranks second in the world and first in Europe on cluster strength, per ECIPE, as cited in VTT's SUPREME announcement. VTT and IQM, the Espoo-based quantum hardware company, have already built Europe's largest 50-qubit superconducting machine, with a public roadmap pointing to 150- and 300-qubit systems. In March 2026, IQM deployed a fourth on-prem 20-qubit system at Aalto University, the Aalto Q20. The infrastructure QScale is meant to feed is concrete, European, and growing.
That said, the project is a launch, not a delivery. A three-year, €5 million photonics program is not going to retire the cryogenic bottleneck by 2029. The million-qubit machine the energy projections reference does not exist, and the nuclear-plant cooling figure is a thought experiment pinned to it. QScale is research, not product, and the consortium's own framing, "experimental stage to industrial-scale applications," is aspirational, not scheduled.
What to watch: QScale's first photonic-microwave control demonstrators, which would be the first proof that the control-cabling bottleneck can be lifted without sacrificing qubit fidelity. The next round of European quantum funding decisions will tell whether this approach gets scaled or stays a Finnish bet.