# Batched Sparse Matrix Multiplication for Accelerating Graph Convolutional Networks

## TL;DR

This paper introduces a novel batched sparse matrix multiplication algorithm optimized for small graphs in GCNs, significantly accelerating training and inference on GPUs by leveraging high parallelism.

## Contribution

It presents the first batched SpMM algorithm tailored for small sparse matrices, improving GPU utilization in GCN applications.

## Key findings

- Achieved up to 1.59x speedup in training
- Achieved up to 1.37x speedup in inference
- Demonstrated effectiveness on NVIDIA Tesla P100 GPU

## Abstract

Graph Convolutional Networks (GCNs) are recently getting much attention in bioinformatics and chemoinformatics as a state-of-the-art machine learning approach with high accuracy. GCNs process convolutional operations along with graph structures, and GPUs are used to process enormous operations including sparse-dense matrix multiplication (SpMM) when the graph structure is expressed as an adjacency matrix with sparse matrix format. However, the SpMM operation on small graph, where the number of nodes is tens or hundreds, hardly exploits high parallelism or compute power of GPU. Therefore, SpMM becomes a bottleneck of training and inference in GCNs applications. In order to improve the performance of GCNs applications, we propose new SpMM algorithm especially for small sparse matrix and Batched SpMM, which exploits high parallelism of GPU by processing multiple SpMM operations with single CUDA kernel. To the best of our knowledge, this is the first work of batched approach for SpMM. We evaluated the performance of the GCNs application on TSUBAME3.0 implementing NVIDIA Tesla P100 GPU, and our batched approach shows significant speedups of up to 1.59x and 1.37x in training and inference, respectively.