Energy-Efficient Cryogenic Neuromorphic Network with Superconducting Memristor

2025-01-13

AI Summary

Conventional computing architectures face fundamental bottlenecks in energy efficiency and scalability for neuromorphic computing. Method: This work proposes a cryogenic brain-inspired computing system based on superconducting memristors (SMs), featuring the first fully integrated superconducting memristive neuromorphic architecture—combining cryogenic spiking neuron circuits, event-driven synaptic topology, and a 23-level pulse coding scheme enabling non-volatile synaptic weight storage. Contribution/Results: Evaluated on the Cart-Pole reinforcement learning control task, the system achieves an average episode length of 5,965 steps (0.02 s/step) over 1,000 trials, with 40% reaching the maximum score of 15,000 steps. It demonstrates significantly reduced power consumption compared to conventional approaches. This work overcomes the longstanding trade-off between energy efficiency and system scale in cryogenic neuromorphic systems and experimentally validates the feasibility of ultra-low-power, real-time dynamic decision-making.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Hardware-aware ML
• Search and Optimization: Learning to Search

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Systems and Infrastructure for Web, Mobile and WoT: Sustainability and carbon-aware systems for Web, mobile, and WoT
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Cryogenic neuromorphic systems, inspired by the brains unparalleled efficiency, present a promising paradigm for next generation computing architectures.This work introduces a fully integrated neuromorphic framework that combines superconducting memristor(SM) based spiking neurons and synapse topologies to achieve a low power neuromorphic network with non volatile synaptic strength.This neurosynaptic framework is validated by implementing the cart pole control task, a dynamic decision making problem requiring real time computation.Through detailed simulations, we demonstrate the network's ability to execute this task with an average fitness of 5965 timesteps across 1000 randomized test episodes, with 40 percent achieving the target fitness of 15,000 timesteps (0.02s per timestep).The system achieves 23 distinct spiking rates across neurons, ensuring efficient information encoding.Our findings establish the potential of SM based cryogenic neuromorphic systems to address the energy and scalability limitations of traditional computing, paving the way for biologically inspired, ultra low power computational frameworks.

Problem

• Energy Efficiency
• Scalability
• Bio-inspired Computing

Innovation

• Superconducting Materials
• Low-Power Neuromorphic Network
• Non-Volatile Synaptic Strength