Helium supply was significantly disrupted at the worst possible time.
In March 2026, as an Iran-Qatar standoff shut down production at Ras Laffan—the world's helium capital, impacting approximately 30% of global supply—quantum computing labs from Zurich to Cambridge scrambled to keep their dilution refrigerators running. Spot prices spiked. Force majeure declarations landed. And somewhere in Lausanne, a six-month-old startup called Rhonexum found itself sitting on what might be the right technology at exactly the right moment.
The company had just closed a $1.0 million pre-seed round (€867.5K, if you're counting in francs) to build chip design tools and control systems that work at temperatures colder than deep space. The timing, announced March 18, wasn't planned around geopolitical chaos. But it didn't hurt.
"Our goal is to miniaturize quantum computing with cryogenic electronics," CEO Vicente Carbon said at the time, though he might as well have added: and maybe solve the helium problem while we're at it.
The Mess Inside the Fridge
Here's the thing about quantum computers that nobody tells you until you're elbow-deep in one: they're held together by an ungodly tangle of wiring.
Each quantum bit—each qubit—needs to be told what to do. Instructions arrive from room-temperature electronics, traveling down meter-long cables into a dilution refrigerator where temperatures hover around 10 millikelvin. That's 0.01 degrees above absolute zero. Colder than the cosmic microwave background. Every wire is a heat leak. Every connection, a potential failure point.
At small scale—a few dozen qubits—the approach is tolerable, if inelegant. At the scale required for practical quantum computing, thousands of qubits or more, it breaks entirely.
A resource-estimation analysis published in January 2026 laid out the math. Heterogeneous control architectures—electronics distributed across room temperature, 4 Kelvin, and millikelvin stages—aren't optional. They're necessary to meet multiplexing and power constraints as systems scale. The paper didn't mince words: the current approach won't work.
Rhonexum emerged from EPFL's AQUA Lab last November with a straightforward pitch. Move the electronics closer to the qubits. Design chips that thrive in the cold rather than merely tolerate it.
The founders—Carbon and Dr. Hung-Chi Han, a former TSMC engineer—built three product lines around that idea. RX-IC offers cryogenic integrated circuits for power management, control, and readout, positioned near the quantum processor itself. RX-MODEL provides what the company calls "cryo-native" design tools, SPICE-compatible models supporting GlobalFoundries 22FD-SOI, STMicroelectronics 28FD-SOI, and TSMC 28nm bulk processes. RX-AMS is an automated test system intended to eliminate manual probing and reduce the thermal-cycle delays that plague development.
The company's modeling approach aims to significantly compress design iteration, using illustrative figures suggesting a reduction from ten runs down to one, cutting development timelines from 80 months to eight. Those figures are illustrative, perhaps aspirational. But the underlying problem they're addressing is real enough.
A Market in Motion
Cryogenic electronics sat at $2.57 billion in valuation last year, according to an SNS Insider report from February. Projections put it at $5.22 billion by 2035—a 7.34% annual growth rate for the 2026-2035 period driven by quantum computing, space instrumentation, and advanced sensing applications. The dilution refrigerator segment alone represents roughly $325 million this year, per late-2025 estimates.
Yet the sector carries visible risk. The helium disruption in March illustrated the volatility and sensitivity of helium-dependent supply chains. Regulatory complexity is rising, too. The U.S. Bureau of Industry and Security added quantum-computing items and certain cryocooling systems to controlled lists in September 2024, complicating cross-border sales. Europe, meanwhile, has allocated €65 million through the Chips Joint Undertaking for quantum chip pilots and is mulling a "Quantum Act" framework that could reshape compliance burdens.
And then there's the technical validation timeline. Quantum computing has a well-earned reputation for promises that slide right. Rhonexum aims to deliver its first industrial-grade systems to early customers by the end of 2026. Whether that holds depends on variables the founders can't entirely control.
Proof Points, Emerging
Still, recent demonstrations suggest the cryo-electronics thesis is gaining traction—and not just in whitepapers.
On March 10, SeeQC published results in Nature Electronics showing a quantum computer controlled entirely by superconducting digital electronics operating at 10 millikelvin, co-located with the qubits. The architecture cut wiring, reduced latency, and minimized heat introduction. Weeks later, at the APS Global Physics Summit, IBM presented data from cryo-CMOS control systems achieving median two-qubit gate fidelities comparable to room-temperature controllers—error rates around 2.3×10^-3, well within the competitive range.
These weren't one-off stunts. IBM has been publishing cryo-CMOS circuit work since at least VLSI 2025, including current references stable across a wide temperature range—5.6 to 100 Kelvin. SemiQon, a Finnish competitor, raised €17.5 million in January 2025 to advance cryo-optimized transistors down to 1 Kelvin, with deliveries expected sometime around now. Delft Circuits, in September 2025, laid out a roadmap to scale planar cryogenic cabling to 1,024 channels per loader by 2027 and 4,096 by 2029.
What Rhonexum offers isn't necessarily breakthrough physics. It's tooling. Infrastructure.
"The lack of cryo-native compact models forces teams to iterate blindly," said Kris Kaczmarek, investment director at QDNL Participations, which led the pre-seed round. "Rhonexum's modeling-led approach can cut fabrication cycles and costs significantly." QDNL participated alongside Venture Kick, with non-dilutive backing from EPFL Startup Launchpad, the Foundation for Innovation and Technology, and the Swiss National Science Foundation.
Whether that's enough differentiation in a crowded field remains an open question.
Beyond Qubits
Carbon frames the mission broadly, and perhaps strategically. The technology targets applications well beyond quantum computing: space instrumentation, deep-space sensors, superconducting detectors (SQUIDs, SNSPDs), trapped-ion and cold-atom systems. YQuantum, a Swiss quantum hardware integrator, announced a collaboration with Rhonexum in early January. "Active cryogenic electronics are a fundamental requirement for scaling quantum computers," Carbon noted at the time, though he was careful to avoid locking the company into a single vertical.
The team—COO Maurice Gaillard, IC designers Georgios Bantemits and Ali Meimandi—is based in Lausanne. Carbon appeared at the Quantum Industry Day Switzerland in October 2025, signaling early ecosystem engagement. The company officially spun out of EPFL less than six months ago. It's young.
That youth cuts both ways. Rhonexum has flexibility, but also limited runway. The pre-seed capital will fund product development toward the late-2026 delivery target. The company hasn't disclosed which customers are queuing up, though its partnership with YQuantum and multi-platform support strategy—superconducting, spin, photonic, trapped-ion, cold-atom—suggest a deliberately broad go-to-market approach.
Convergence or Hype?
The broader market seems to be converging on hybrid architectures. Four-Kelvin cryo-CMOS handling data conversion and multiplexing. Selective millikelvin superconducting logic for ultra-low-latency tasks. Room-temperature controllers managing orchestration. The January resource-estimation analysis anticipates this stack becoming standard for medium-scale systems—10^3 to 10^4 physical qubits—before the decade closes. Vendors demonstrating parity-level gate fidelities with cryogenic control in 2026 lend weight to that forecast.
Perhaps the helium crisis accelerates adoption. Lower heat budgets mean smaller refrigerators, less helium consumption, reduced vulnerability to geopolitical supply shocks. Or perhaps it doesn't matter. Quantum timelines slip. Markets pivot. The cryo-electronics sector serves more than just quantum computers; it'll survive even if fault-tolerant quantum computing remains perpetually five years away.
Either way, Rhonexum is betting that the path to scale runs through the cold. And that the real bottleneck isn't the wiring itself, but the absence of tools to design what replaces it.
That's a narrower bet than it sounds. Tool vendors don't win by being first. They win by being indispensable. Whether Rhonexum can thread that needle—deliver industrial-grade systems on time, build sticky customer relationships, fend off better-funded competitors—will become clearer by year's end.
For now, the helium is still scarce. The refrigerators are still running. And somewhere in a Lausanne lab, a handful of engineers are trying to figure out how to make electronics work where physics says they probably shouldn't.
