Indian Researchers Develop Molecular Memristor for Neuromorphic Computing

2026-04-22 · Source: Big Data & AI News - EE Times · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Expert, medium

Summary

Researchers at the Indian Institute of Science (IISc) Bangalore's Centre for Nano Science and Engineering (CeNSE) have developed a molecular memristor based on a ruthenium compound, achieving 14-bit analog resolution and 4.1 TOPS/W energy efficiency. This innovation enables linear weight updates and improved energy efficiency for neuromorphic computing, particularly in signal processing and edge AI applications. The ruthenium complex utilizes azo-aromatic ligands, exhibiting zeroth-order kinetics for predictable, constant-rate switching and linear, symmetric conductance updates across four orders of magnitude. The team is incubating a startup to commercialize a prototype chip, expected within the next year, designed on a 22-nm process using a crossbar architecture for in-memory computation. This technology addresses the energy limitations of current AI systems, where DRAM access consumes 94% of CNN energy.

Key takeaway

For research scientists developing energy-efficient AI hardware, this molecular memristor technology presents a viable path to overcome power and latency constraints in edge AI. You should consider exploring analog computation and brain-inspired architectures, particularly those compatible with existing silicon manufacturing processes, to achieve significant gains in TOPS/W and reduce reliance on conventional GPU-based systems. The deterministic nature and high endurance of these molecular devices offer a promising foundation for future neuromorphic chip designs.

Key insights

Molecular memristors offer high-resolution, energy-efficient neuromorphic computing for edge AI by performing in-memory analog computation.

Principles

• Analog computation reduces energy consumption.
• Zeroth-order kinetics ensures predictable switching.
• Crossbar architecture enables parallel matrix operations.

Method

The method involves engineering a ruthenium-based molecular memristor with azo-aromatic ligands, arranging millions in a crossbar grid for in-memory analog computation, and integrating it into a mixed-signal SoC.

In practice

• Target 8-bit resolution for AI inference.
• Integrate into standard CMOS flows.
• Explore iron/cobalt alternatives for materials.

Topics

• Molecular Memristor
• Neuromorphic Computing
• Edge AI
• Ruthenium Compound
• Crossbar Architecture

Best for: Research Scientist, AI Scientist, AI Hardware Engineer, Director of AI/ML

Related on AIssential

• This new chip survives 1300°F (700°C) and could change AI forever — A new memristor design withstands 700°C, enabling extreme environment electronics and efficient AI computation.
• Algorithm-hardware co-design of neuromorphic networks with dual memory pathways — A dual memory pathway architecture and co-designed hardware significantly enhance neuromorphic network efficiency and temporal processing.
• Brain-inspired phototransistor could cut AI energy use by sensing and storing data — Brain-inspired phototransistor integrates sensing and memory to reduce AI energy consumption.
• Positional Encoding in the Context of Memristor-Based Analog Computation for Automatic Speech Recognition — Large positional encoding outputs degrade memristor-based ASR computation, mitigated by ADC adjustment or transformation removal.
• Benchmarking the Energy Cost of Assurance in Neuromorphic Edge Robotics — Neuromorphic defenses can enhance AI robustness while simultaneously reducing energy consumption due to induced sparsity.
• Brain-inspired chip runs near absolute zero and could transform quantum computing — A brain-inspired SiC chip operates near absolute zero, enabling integrated quantum computing control and deep space applications.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Big Data & AI News - EE Times.