KNN Technology Desk | Joseph — A research team supported by the U.S. National Science Foundation has demonstrated a nanoscale device that can both process an electrical signal and retain a temporary memory of earlier inputs. The agency described the findings on October 29, 2026, after their publication in the scientific journal Science Advances. The work is a laboratory demonstration, not a consumer computer chip ready for commercial production.
How a soft polymer becomes a tiny computing component
Researchers placed an extremely thin layer of polydimethylsiloxane, a soft polymer, between two electrodes. As voltage is applied, the electrodes move closer together and compress the material. The polymer acts like a spring, allowing the movement to reverse rather than permanently sticking the metal parts together.
That movement affects the electrical current. As the voltage builds, the device can produce a stronger response after reaching a threshold. Then, as the polymer relaxes over time, its response depends partly on its previous state. This is the short-lived memory that lets the same structure combine simple processing and information storage.
Why researchers compare it with a neuron
Biological neurons integrate signals and respond when their activity crosses a threshold. The research device mimics selected aspects of that pattern through mechanical motion and electrical changes. It is not a living neuron, and it does not demonstrate consciousness, thought or the full complexity of a brain. The analogy helps researchers design computing components in which memory and processing are less separated than they are in many conventional systems.
NSF says possible future applications could include low-power wearable sensors, environmental monitors and systems that process touch signals. Those are potential uses rather than demonstrated finished products. The agency says the project was supported through its molecularly optimized networks center and national nanotechnology infrastructure network.
What has not been established
Energy use, durability, manufacturability and integration into large circuits will need additional validation before the approach can compete with conventional electronics. The report should not be interpreted as proof that brain-like computers have replaced semiconductor processors.
Primary sources: NSF research summary, October 29, 2026 and the peer-reviewed study in Science Advances.
