NAND-SPIN-Based Processing-in-MRAM Architecture for Convolutional Neural Network Acceleration

TL;DR

This paper introduces a NAND-SPIN-based processing-in-memory architecture designed to accelerate convolutional neural networks, significantly improving speed and energy efficiency by leveraging in-memory computation and optimized data mapping.

Contribution

It presents a novel NAND-SPIN PIM architecture with a data mapping scheme that enhances parallelism and reduces data movement for CNN acceleration.

Findings

Achieves approximately 2.6x speedup over existing solutions.

Provides about 1.4x energy efficiency improvement.

Demonstrates effective in-memory CNN processing with NAND-SPIN technology.

Abstract

The performance and efficiency of running large-scale datasets on traditional computing systems exhibit critical bottlenecks due to the existing "power wall" and "memory wall" problems. To resolve those problems, processing-in-memory (PIM) architectures are developed to bring computation logic in or near memory to alleviate the bandwidth limitations during data transmission. NAND-like spintronics memory (NAND-SPIN) is one kind of promising magnetoresistive random-access memory (MRAM) with low write energy and high integration density, and it can be employed to perform efficient in-memory computation operations. In this work, we propose a NAND-SPIN-based PIM architecture for efficient convolutional neural network (CNN) acceleration. A straightforward data mapping scheme is exploited to improve the parallelism while reducing data movements. Benefiting from the excellent characteristics of NAND-SPIN and in-memory processing architecture, experimental results show that the proposed approach can achieve $\sim$2.6$\times$ speedup and $\sim$1.4$\times$ improvement in energy efficiency over state-of-the-art PIM solutions.