AutoGMap: Learning to Map Large-scale Sparse Graphs on Memristive Crossbars

TL;DR

This paper introduces a reinforcement learning-based dynamic sparsity-aware mapping scheme for large-scale sparse graph computation on memristive crossbars, significantly reducing resource usage and improving efficiency.

Contribution

It proposes a novel RL-based method for dynamic, sparsity-aware mapping of large-scale graphs on memristive crossbars, outperforming static and coarse-grained approaches.

Findings

Achieves 43% area of original matrix on small-scale data

Reduces area to 22.5% on qh882 and 17.1% on qh1484 datasets

Demonstrates potential extension to other PIM architectures

Abstract

The sparse representation of graphs has shown great potential for accelerating the computation of graph applications (e.g., Social Networks, Knowledge Graphs) on traditional computing architectures (CPU, GPU, or TPU). But the exploration of large-scale sparse graph computing on processing-in-memory (PIM) platforms (typically with memristive crossbars) is still in its infancy. To implement the computation or storage of large-scale or batch graphs on memristive crossbars, a natural assumption is that a large-scale crossbar is demanded, but with low utilization. Some recent works question this assumption, to avoid the waste of storage and computational resource, the fixed-size or progressively scheduled ''block partition'' schemes are proposed. However, these methods are coarse-grained or static, and are not effectively sparsity-aware. This work proposes the dynamic sparsity-aware mapping scheme generating method that models the problem with a sequential decision-making model, and optimizes it by reinforcement learning (RL) algorithm (REINFORCE). Our generating model (LSTM, combined with the dynamic-fill scheme) generates remarkable mapping performance on a small-scale graph/matrix data (complete mapping costs 43% area of the original matrix) and two large-scale matrix data (costing 22.5% area on qh882 and 17.1% area on qh1484). Our method may be extended to sparse graph computing on other PIM architectures, not limited to the memristive device-based platforms.