Can Superfluid Helium Help Build Better Quantum Computers?

Quantum computing has a hardware problem: the same interactions that let researchers control quantum information can also disturb it. Electrical fluctuations, material defects, and unwanted energy exchanges can introduce errors before a calculation finishes. Building useful machines requires balancing access to fragile quantum states with protection from their surroundings.

A paper by Priya Sharma, Jens Koch, and Eran Ginossar, published in npj Quantum Information, proposes an unusual approach: a microscopic device that stores quantum information in the motion of superfluid helium-3.¹ Rather than relying on electrical current as its central operating mechanism, the proposed device combines neutral fluid flow with a flexible mechanical plate.

The attraction is reduced exposure to certain electrical noise sources. The challenge is that this potential advantage comes with demanding cooling, fabrication, and control requirements. The study establishes a theoretical route toward a new qubit, while leaving experimental performance and practical scaling unresolved.

How a Superfluid Helium Qubit Would Work

Helium-3 becomes superfluid at extremely low temperatures, allowing a component of the liquid to flow without ordinary viscous resistance. Its collective behavior shares important features with superconductivity, including a version of the Josephson effect, which connects flow through a weak link to the quantum phase difference across it.

The proposed Superfluid Helium Oscillator Quantum device, abbreviated SHOQ, contains a small cell connected to a surrounding helium reservoir through a nanoscale aperture. An elastic plate forms the cell’s lid. As helium moves through the aperture, pressure changes inside the cell move the plate.

The fluid and plate together form an oscillator. Under suitable conditions, its energy is quantized, meaning it occupies discrete energy levels rather than any arbitrary energy. Two of those levels could represent the states used to encode a quantum bit.

However, discrete levels alone are insufficient. Their spacing must also differ enough for researchers to address one transition without accidentally exciting another. This uneven spacing, called anharmonicity, is essential to making an oscillator behave as a controllable qubit. The authors show how aperture dimensions, plate size, and mechanical stiffness could provide that necessary nonlinearity.

Why Charge Neutrality Matters for Quantum Computing

In superconducting qubits, quantum information is encoded in electrical circuit behavior. The SHOQ proposal instead uses the collective motion of electrically neutral helium. The authors suggest that charge and flux noise may therefore be less problematic for this architecture.

That distinction is potentially valuable, but it does not establish immunity to noise. Mechanical components, energy leakage through the aperture, and the equipment used to operate the device introduce other ways for information to degrade.

The broader lesson is that changing qubit physics changes the engineering problem. A platform can reduce one vulnerability while creating a different bottleneck. Securities.io’s comparison of quantum hardware architectures provides context for why competing approaches require different infrastructure and manufacturing strategies.

Charge neutrality also needs careful interpretation. This proposal concerns a collective state of liquid helium, rather than individually trapped neutral atoms. Both involve neutral matter, but their mechanisms, temperatures, and control requirements differ substantially.

What the Helium Qubit Calculations Show

The researchers evaluate six example designs. Their calculations suggest that suitable devices could retain quantum information long enough to support qubit operations, assuming the modeled dissipation mechanisms adequately describe the finished hardware.

The table below compares three designs directly from the paper. The lifetime estimates reflect dissipation from single quasiparticle tunneling, which the authors identify as the dominant low-temperature loss mechanism in their analysis.

These are theoretical estimates, rather than measured coherence times. In particular, an estimated energy-relaxation lifetime should not be treated as a complete experimental demonstration of how long a qubit preserves phase information.

The authors identify design #5 as especially promising. Nevertheless, longer lifetime alone does not determine usefulness. Gate speed, leakage into unwanted levels, initialization, measurement accuracy, and coupling between devices all affect how much reliable computation a platform can perform.

Extreme Cooling Is the Central Engineering Tradeoff

The temperature dependence is striking. For design #5, lowering the temperature from 0.4 to 0.3 millikelvin increases the calculated lifetime from 0.64 to 25.96 milliseconds, approximately fortyfold.

Those temperatures equal just 400 and 300 microkelvin above absolute zero. The paper notes that temperatures around 0.4 millikelvin have been reached in previous helium-3 experiments. That supports physical feasibility, but operating a controlled quantum device under those conditions remains a demanding task.

This creates a useful distinction between device size and system size. A qubit with a plate only a few micrometres across does not automatically produce a compact computer. Its surrounding cooling equipment, connections, shielding, and measurement apparatus may dominate the installation.

A practical assessment would therefore ask how many reliable operations the complete system can deliver for its cooling burden and operating complexity. If lower electrical noise requires substantially more difficult refrigeration, the advantage must be large enough to justify that infrastructure.

Temperature also affects initialization. For one evaluated design, the paper calculates a 10.3% thermal occupation of the first excited state at 0.3 millikelvin. Simply cooling the device would therefore not guarantee that every operation starts in the desired state.

Controlling a Neutral Qubit Still Requires an Interface

A protected quantum state is useful only if researchers can prepare, manipulate, and measure it. The paper illustrates a hybrid approach involving the SHOQ device, a superconducting transmon qubit, and a microwave cavity.

The mechanical plate’s motion would change an electrical capacitance, allowing the helium oscillator to interact with the superconducting circuit. That circuit could help translate between the neutral fluid motion and conventional electronic control and measurement.

This arrangement reveals a second tradeoff. The central information-carrying mechanism may avoid some electrical vulnerabilities, while its control interface reconnects it to electrical hardware. The performance of the combined system must therefore be measured, rather than inferred solely from the isolated helium oscillator.

The paper leaves detailed gate architecture and operation for future work. Important experimental milestones include:

• Resolving the predicted energy levels and their uneven spacing.
• Demonstrating reliable state preparation, control, and measurement.
• Measuring relaxation and phase coherence in fabricated devices.
• Coupling multiple devices while preserving useful performance.

Quantum Fabrication Offers an Investment Connection

These requirements connect the research to an existing industrial challenge: manufacturing delicate quantum devices consistently. A nanoscale aperture or elastic membrane must behave predictably after fabrication, assembly, and cooling. Variation between nominally identical components could complicate control and reduce yield.

For investors interested in this enabling technology, Oxford Instruments offers a relevant public company to examine. Its plasma processing equipment supports precise etching and deposition, including tools already being purchased for quantum device fabrication.

In May 2026, Oxford Instruments announced that Rigetti Computing (RGTI ) had purchased its PlasmaPro 100 Cobra system for atomic layer etching for quantum devices. The announcement describes a low-damage process intended to support Rigetti’s expanded fabrication capabilities.

Separately, a May announcement outlined an atomic-scale fabrication partnership with NYU Nanofab. Together, these developments illustrate demand for manufacturing tools across quantum research and development.

This is an indirect investment connection. The helium paper does not identify Oxford Instruments as a participant or establish that its equipment will manufacture SHOQ devices. The company’s 2026 full-year results also confirm the disposal of its NanoScience business, making fabrication the appropriate focus rather than its former cryogenics portfolio.

From a Qubit Proposal to Useful Quantum Hardware

The helium concept occupies an early stage of development. Its calculations must become reproducible measurements before comparisons with established platforms carry much weight.

Meanwhile, other architectures are progressing toward customer installations, as illustrated by the announced quantum computer acquisition by Florida International University. That difference in maturity matters when evaluating a new proposal’s commercial significance.

The value of this research is the alternative it opens: quantum information encoded in a charge-neutral condensed-matter system. Its eventual success will depend on whether that physical advantage survives the cooling, fabrication, and control requirements of a complete machine. Demonstrating that balance would be more consequential than an impressive lifetime estimate alone.

References:

¹ Sharma, P., Koch, J., & Ginossar, E. (2026). Towards a micromechanical qubit based on quantized oscillations in superfluid helium. npj Quantum Information. Advance online publication. https://doi.org/10.1038/s41534-026-01355-3