GNN-PIM: A Processing-in-Memory Architecture for Graph Neural Networks

Wang, Zhao; Guan, Yijin; Sun, Guangyu; Niu, Dimin; Wang, Yuhao; Zheng, Hongzhong; Han, Yinhe

doi:10.1007/978-981-15-8135-9_6

Cited by 9 publications

(4 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the software side, several libraries have been proposed to improve the support for GNNs and handle its multiple variants, being the extensions of popular libraries such as PyTorch or Tensorflow (TF) [47], [162], [163] clear examples. On the hardware side, new accelerator architectures have been surfacing recently [40], [42], [117], [164] that attempt to deal with the flexibility and scalability challenges of GNNs. In the next subsections, we provide an exhaustive overview of existing techniques.…”

Section: Algorithm Aggregation (A)mentioning

confidence: 99%

Computing Graph Neural Networks: A Survey from Algorithms to Accelerators

et al. 2021

View full text Add to dashboard Cite

Graph Neural Networks (GNNs) have exploded onto the machine learning scene in recent years owing to their capability to model and learn from graph-structured data. Such an ability has strong implications in a wide variety of fields whose data are inherently relational, for which conventional neural networks do not perform well. Indeed, as recent reviews can attest, research in the area of GNNs has grown rapidly and has lead to the development of a variety of GNN algorithm variants as well as to the exploration of ground-breaking applications in chemistry, neurology, electronics, or communication networks, among others. At the current stage research, however, the efficient processing of GNNs is still an open challenge for several reasons. Besides of their novelty, GNNs are hard to compute due to their dependence on the input graph, their combination of dense and very sparse operations, or the need to scale to huge graphs in some applications. In this context, this article aims to make two main contributions. On the one hand, a review of the field of GNNs is presented from the perspective of computing. This includes a brief tutorial on the GNN fundamentals, an overview of the evolution of the field in the last decade, and a summary of operations carried out in the multiple phases of different GNN algorithm variants. On the other hand, an in-depth analysis of current software and hardware acceleration schemes is provided, from which a hardware-software, graph-aware, and communication-centric vision for GNN accelerators is distilled.

show abstract

Section: Algorithm Aggregation (A)mentioning

confidence: 99%

Computing Graph Neural Networks: A Survey from Algorithms to Accelerators

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The high density and complexity of GCNs make the on-chip communication for IMC-based accelerators even more critical. Authors in [30,31] proposed an IMC-based accelerator for GCN. However, these technique does not address the issue of the on-chip communication performance of GCN accelerators.…”

Section: Related Workmentioning

confidence: 99%

COIN: Communication-Aware In-Memory Acceleration for Graph Convolutional Networks

Mandal,

Krishnan,

Goksoy

et al. 2022

Preprint

View full text Add to dashboard Cite

Graph convolutional networks (GCNs) have shown remarkable learning capabilities when processing graphstructured data found inherently in many application areas. GCNs distribute the outputs of neural networks embedded in each vertex over multiple iterations to take advantage of the relations captured by the underlying graphs. Consequently, they incur a significant amount of computation and irregular communication overheads, which call for GCN-specific hardware accelerators. To this end, this paper presents a communicationaware in-memory computing architecture (COIN) for GCN hardware acceleration. Besides accelerating the computation using custom compute elements (CE) and in-memory computing, COIN aims at minimizing the intra-and inter-CE communication in GCN operations to optimize the performance and energy efficiency. Experimental evaluations with widely used datasets show up to 105× improvement in energy consumption compared to state-of-the-art GCN accelerator.

show abstract

“…Recently, plenty of GCN/GNN accelerators have been presented [21]- [29], [80] for efficient GCN inference. As far as we know, GraphACT [54] is the only GCN training accelerator.…”

Section: Related Work a Gcn Accelerationmentioning

confidence: 99%

“…Various GCN accelerators have been proposed to accelerate GCN inference [21]- [29]. However, existing accelerators and GPU/CPU platforms do not easily support GCN training at scale because GCN training tasks have many peculiar challenges.…”

Section: Introductionmentioning

confidence: 99%

GCNear: A Hybrid Architecture for Efficient GCN Training with Near-Memory Processing

Zhu¹,

Liu²,

Wei³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Recently, Graph Convolutional Networks (GCNs) have become state-of-the-art algorithms for analyzing noneuclidean graph data. However, it is challenging to realize efficient GCN training, especially on large graphs. The reasons are many-folded: 1) GCN training incurs a substantial memory footprint. Full-batch training on large graphs even requires hundreds to thousands of gigabytes of memory to buffer the intermediate data for back-propagation. 2) GCN training involves both memory-intensive data reduction and computation-intensive features/gradients update operations. Such a heterogeneous nature challenges current CPU/GPU platforms. 3) The irregularity of graphs and the complex training dataflow jointly increase the difficulty of improving a GCN training system's efficiency.This paper presents GCNear, a hybrid architecture to tackle these challenges. Specifically, GCNear adopts a DIMM-based memory system to provide easy-to-scale memory capacity. To match the heterogeneous nature, we categorize GCN training operations as memory-intensive Reduce and computation-intensive Update operations. We then offload Reduce operations to on-DIMM NMEs, making full use of the high aggregated local bandwidth. We adopt a CAE with sufficient computation capacity to process Update operations. We further propose several optimization strategies to deal with the irregularity of GCN tasks and improve GCNear's performance. Comprehensive evaluations on twelve GCN training tasks demonstrate that GCNear achieves 24.8× / 2.2× geomean speedup and 61.9× / 6.4× (geomean) higher energy efficiency compared to Xeon E5-2698-v4 CPU and NVIDIA V100 GPU. To deal with deep GCN models and the ever-increasing graph scale, we also propose a Multi-GCNear system. Compared to state-of-the-art Roc and DistGNN systems, Multi-GCNear achieves up to 2.1× and 3.1× higher training speed, respectively.

show abstract

GNN-PIM: A Processing-in-Memory Architecture for Graph Neural Networks

Cited by 9 publications

References 26 publications

Computing Graph Neural Networks: A Survey from Algorithms to Accelerators

Computing Graph Neural Networks: A Survey from Algorithms to Accelerators

COIN: Communication-Aware In-Memory Acceleration for Graph Convolutional Networks

GCNear: A Hybrid Architecture for Efficient GCN Training with Near-Memory Processing

Contact Info

Product

Resources

About