Self-heating electrochemical memory for high-precision analog computing
🤖 AI Summary
AI hardware faces severe energy-efficiency bottlenecks, and conventional analog in-memory computing (AIMC) suffers from fundamental trade-offs among precision, linearity, and energy efficiency.
Method: This work proposes a self-heating gate-controlled memristor leveraging electrochemical modulation of oxygen vacancies in oxide thin films. A single gate electrode simultaneously enables Joule heating and electrochemical driving, dramatically enhancing oxygen vacancy mobility.
Contribution/Results: The device achieves a 9-decade dynamic resistance range, ultra-low-voltage programming (<2 V), 15-ns write speed, and over 3,000 stable analog states. It exhibits six-decade linear current–voltage response and deterministic high-precision writing. Integrated into a low-power analog in-memory computing architecture, it enables up to 100× energy-efficiency improvement over state-of-the-art digital accelerators, advancing practical high-precision, wide-dynamic-range analog compute-in-memory systems.
Application Category
📝 Abstract
Artificial intelligence (AI) is pushing the limits of digital computing to such an extent that, if current trends were to continue, global energy consumption from computation alone would eclipse all other forms of energy within the next two decades. One promising approach to reduce energy consumption and to increase computational speed is in-memory analog computing. However, analog computing necessitates a fundamental rethinking of computation at the material level, where information is stored as continuously variable physical observables. This shift introduces challenges related to the precision, dynamic range, and reliability of analog devices - issues that have hindered the development of existing memory technology for use in analog computers. Here, we address these issues in the context of memory which stores information as resistance. Our approach utilizes an electrochemical cell to tune the bulk oxygen-vacancy concentration within a metal oxide film. Through leveraging the gate contact as both a heater and source of electrochemical currents, kinetic barriers are overcome to enable a dynamic range of nine decades of analog tunable resistance, more than 3,000 available states, and programming with voltages less than 2 V. Furthermore, we demonstrate deterministic write operations with high precision, current-voltage linearity across six decades, and programming speeds as fast as 15 ns. These characteristics pave the way toward low-power analog computers with potential to improve AI efficiency by orders of magnitude.
Problem
Research questions and friction points this paper is trying to address.
Reducing energy consumption in AI computing with analog memory
Overcoming precision and reliability issues in analog devices
Achieving high dynamic range and fast programming in resistive memory
Innovation
Methods, ideas, or system contributions that make the work stand out.
Self-heating electrochemical memory for analog computing
Nine-decade dynamic range with 3,000 resistance states
High-precision deterministic write operations at 15ns
🔎 Similar Papers
No similar papers found.
