A tiny vortex in a superconducting material makes it possible to manipulate a bit with an energy cost close to a fundamental limit of physics.
A classical bit can take two states, denoted 0 and 1. Here, the researchers do not use an electrical charge stored in a transistor. They exploit a superconducting vortex, a tiny region around which a current circulates in the material. Its position directly represents one of the two bit states.
Experimental device using a superconducting vortex confined in a nanometric trapping potential.
Credit: Fridman et al., Nature Communications (2026), CC BY 4.0 license.
To maintain this vortex, the team created a nanometric trap with two stable positions. A force applied to the system can move the vortex from one position to the other: 0 or 1. The researchers studied the force required for these movements to determine the energy needed to change a 0 into a 1 and vice versa.
The experiment operates at 4.2 K, or approximately -268.95 °C. At this temperature, the material is superconducting: current can flow through it without the usual electrical resistance.
The energy challenge arises when a bit must be changed. Physics imposes a minimum energy cost for this operation, more precisely for erasure: this cost is called the Landauer limit. It depends on temperature. At 4.2 K, this limit is approximately 4 × 10-23 joules per bit, an extremely small amount on the scale of an electronic device.
The researchers did not, however, directly measure an operation following the Landauer protocol. Their experimental device is incompatible with this precise protocol. They therefore used numerical simulations to evaluate the energy of a suitable manipulation. The result is less than a factor of ten above the theoretical limit.
The research study therefore does not present a complete computer memory consuming almost the absolute minimum. It shows that a bit based on the position of a superconducting vortex can, in the model studied, come considerably close to it. Its operation at very low temperatures also limits this approach to systems that are already cooled; otherwise, the energy benefit obviously loses all meaning.
The authors are considering, in particular, integration with cryogenic quantum circuits, which also operate at low temperatures. They also mention a possible use of these vortices in certain quantum computing architectures. The next steps will have to transform this isolated bit into controllable elements that can be integrated into larger circuits.