Visco-Bits: temporal-coding-inspired viscoelastic metamaterial for mechanical memory expansion

Area of Science:

• * Materials Science and Engineering
• * Computational Science and Engineering
• * Physics of Information

Background:

• * Traditional digital computing faces scalability challenges, particularly with von Neumann architectures.
• * Miniaturization alone is insufficient for enhancing information density in intelligent structures.
• * Neuromorphic and in-memory computing offer alternative strategies to overcome these limitations.

Purpose of the Study:

• * To explore mechanical metamaterials for memory augmentation and in-material computation.
• * To leverage viscoelastic properties for temporal neural coding and information encoding/decoding.
• * To investigate applications in energy-limited environments or where embedded electronics are infeasible.

Main Methods:

• * Utilized an asymmetric bistable viscoelastic mechano-bit (Visco-Bit) for information encoding.
• * Developed a decoding protocol based on the global stiffness of the mechanical storage system.
• * Derived memory capacity limits and identified a geometric stiffness design rule for elastic mechano-bits.

Main Results:

• * Demonstrated transient memory augmentation via neuron-inspired viscoelastic mechano-bits (Visco-Bits).
• * Showcased temporal stiffness modulation leading to augmented mechanical memory storage through non-abelian, path-dependent behavior.
• * Exceeded conventional n-bit digital storage limits in information entropy and performed logic gate operations.

Conclusions:

• * Viscoelastic mechano-bits offer a viable pathway for advanced mechanical computing and memory augmentation.
• * The proposed abstraction extends to various physical systems capable of temporal property modulation.
• * This research paves the way for intelligent structures with enhanced computational capabilities in resource-constrained settings.