Beyond the Device: Signal, Scale, and System Bottlenecks in Neuromorphic Computing

Novel neuromorphic hardware has demonstrated brain-like capabilities with real potential to enhance computing tasks, from pattern recognition to adaptive control. As these devices move from laboratory demonstrations toward practical deployment, a distinct challenge has emerged: the greatest barrier to realizing their potential no longer lies solely in the device itself, but at the interface between that hardware and the rest of the system. Converting signals between analog and digital domains can carry a greater cost than the underlying neuromorphic operation, and as systems scale, coordinating updates across distributed hardware elements introduces bottlenecks that limit overall performance. Addressing this interface, rather than the device in isolation, is now central to progress in the field.This Research Topic aims to bring together contributions addressing the interface and scaling challenges that stand between promising neuromorphic devices and deployable, high-performance systems. We seek to clarify where the field’s real bottlenecks lie: whether at the circuit level, in signal conversion and interconnect design, or at the system level, in memory access, communication, and update propagation across distributed hardware. By convening work spanning devices, circuits, and architectures in a single collection, we hope to surface design principles and trade-offs that are difficult to see from within any one layer alone, and to chart practical paths toward neuromorphic systems that scale efficiently in both energy and performance.We welcome contributions on:- Circuit and device-level solutions for analog-digital signal conversion in neuromorphic hardware- Architectures for managing communication, memory access, and update propagation at scale- Methods for reducing the energy and latency cost of hardware-system interfaces- Studies connecting device physics to system-level performance metrics- Perspectives on scalable, efficient neuromorphic system designWe invite Original Research, Review, and Perspective articles. Topic editors: William Borders, Gina Adam, Ramtin Zand Source: https://www.frontiersin.org/research-topics/86003/beyond-the-device-signal-scale-and-system-bottlenecks-in-neuromorphic-computing