About this document
Neuromorphic Computing: On-chip Interconnects by hspark is a document available to read on EtoBox.
The document discusses a neuromorphic computing system that integrates computation into memory using non-volatile resistive memory (ReRAM) to minimize off-chip memory access. It models and analyzes the effects of interconnects on the accuracy and power consumption of deep neural networks (DNN) within this architecture. The study highlights the importance of on-chip interconnects and their impact on system performance, addressing challenges such as IR drop and crosstalk.
- Author
- hspark
- Language
- EN