HePGA: A Heterogeneous Processing-in-Memory based GNN Training Accelerator

TL;DR

HePGA is a novel 3D heterogeneous PIM-based accelerator that optimizes GNN training by leveraging multiple PIM devices, achieving significant improvements in energy and compute efficiency without losing accuracy.

Contribution

This work introduces HePGA, the first 3D heterogeneous PIM architecture tailored for GNN training, combining various PIM devices for enhanced performance.

Findings

HePGA outperforms existing PIM architectures by up to 3.8x in energy efficiency.

HePGA achieves up to 6.8x better compute efficiency.

HePGA maintains GNN prediction accuracy while improving efficiency.

Abstract

Processing-In-Memory (PIM) architectures offer a promising approach to accelerate Graph Neural Network (GNN) training and inference. However, various PIM devices such as ReRAM, FeFET, PCM, MRAM, and SRAM exist, with each device offering unique trade-offs in terms of power, latency, area, and non-idealities. A heterogeneous manycore architecture enabled by 3D integration can combine multiple PIM devices on a single platform, to enable energy-efficient and high-performance GNN training. In this work, we propose a 3D heterogeneous PIM-based accelerator for GNN training referred to as HePGA. We leverage the unique characteristics of GNN layers and associated computing kernels to optimize their mapping on to different PIM devices as well as planar tiers. Our experimental analysis shows that HePGA outperforms existing PIM-based architectures by up to 3.8x and 6.8x in energy-efficiency (TOPS/W) and compute efficiency (TOPS/mm2) respectively, without sacrificing the GNN prediction accuracy. Finally, we demonstrate the applicability of HePGA to accelerate inferencing of emerging transformer models.