LGARETApr 16, 2019

Processing-In-Memory Acceleration of Convolutional Neural Networks for Energy-Efficiency, and Power-Intermittency Resilience

arXiv:1904.07864v118 citations
Originality Incremental advance
AI Analysis

This work addresses energy and resilience challenges for battery-less IoT nodes, representing a domain-specific advancement with incremental improvements in hardware acceleration.

The paper tackles the problem of energy efficiency and power-intermittency resilience in convolutional neural networks (CNNs) by implementing a bit-wise in-memory accelerator using SOT-MRAM with a novel AND-Accumulation method, achieving up to ~9.7x higher energy-efficiency and 13.5x speedup over CMOS-only approaches while maintaining comparable inference accuracy.

Herein, a bit-wise Convolutional Neural Network (CNN) in-memory accelerator is implemented using Spin-Orbit Torque Magnetic Random Access Memory (SOT-MRAM) computational sub-arrays. It utilizes a novel AND-Accumulation method capable of significantly-reduced energy consumption within convolutional layers and performs various low bit-width CNN inference operations entirely within MRAM. Power-intermittence resiliency is also enhanced by retaining the partial state information needed to maintain computational forward-progress, which is advantageous for battery-less IoT nodes. Simulation results indicate $\sim$5.4$\times$ higher energy-efficiency and 9$\times$ speedup over ReRAM-based acceleration, or roughly $\sim$9.7$\times$ higher energy-efficiency and 13.5$\times$ speedup over recent CMOS-only approaches, while maintaining inference accuracy comparable to baseline designs.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes