Researchers at the University of Surrey have published a design for a new kind of qubit for quantum computing. Today’s leading machines often use superconducting circuits, which are electrical loops cooled until they carry current with no resistance. Those circuits are easily disturbed by electromagnetic noise and by leftover electric charge. Even small disturbances can scramble the stored information and make large machines hard to keep accurate.
The Surrey design, published in npj Quantum Information, is based on superfluid helium-3. Helium-3 is a rare form of helium. When cooled far enough, it becomes a superfluid, a liquid that flows with no friction. The proposed device is called SHOQ, for Superfluid Helium Oscillator Quantum. It would use this charge-neutral liquid inside a microfluidic chip, a tiny channel system that handles very small volumes of fluid. Because the working material carries no electric charge, the authors argue it should be naturally less sensitive to some kinds of electrical noise. Their calculations suggest error rates about 100 times lower than those of typical superconducting qubits. This is presented as the first reported design for a qubit based on a superfluid.
Still a design on paper
The separate ingredients were already known. What is new is putting them together in one microfluidic device and working out the numbers needed for it to act as a qubit. The mathematics indicates that it should work, and that the next step is to build a prototype and test the predictions. The researchers note that one kind of qubit may not have to do every job. A later version of SHOQ might serve as quantum memory - a place to store quantum information while other hardware does the calculations. The paper also describes how the device could be linked to existing superconducting circuits rather than replace them.
SHOQ would still need extremely low temperatures, but those conditions have already been reached in helium-3 experiments. A commercialisation fellowship is supporting the attempt to build a first device. Until that prototype exists, the low error rates remain a prediction, not a measured result.