On Efficient Scaling of GNNs via IO-Aware Layers Implementations

• URL: http://arxiv.org/abs/2605.31500v1
• Date: Fri, 29 May 2026 16:22:45 GMT
• Title: On Efficient Scaling of GNNs via IO-Aware Layers Implementations
• Abstract summary: Graph Neural Networks (GNNs) are bottlenecked by irregular memory access.<n>We show that widely used layers fall into three kernel families: SpMM-based convolutions, reduction-based aggregations, and attention-based layers.<n>For each family, we develop kernels that reduce data movement, improve locality, and remain robust across realistic graphs.
• Score: 1.0262304700896199
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Graph Neural Networks (GNNs) are bottlenecked by sparse, irregular memory access. Popular frameworks such as DGL and PyTorch Geometric support general message passing, but complex layers often materialize edge-wise intermediates, increasing memory traffic and limiting scalability on large graphs. We take an I/O- and arithmetic-intensity--centric view and show that widely used layers fall into three kernel families: SpMM-based convolutions, reduction-based aggregations, and attention-based layers (GATv2/Graph Transformer). For each family, we develop GPU kernels that reduce data movement, improve locality, and remain robust across realistic graphs. We also study graph reordering and find that its impact depends on the kernel mapping: it benefits neighbor-parallel (gather-dominated) kernels more consistently than feature-parallel designs. Empirically, our fused attention kernels reach up to $\textbf{3.9}\times$ speedup for Graph Transformer (median $\textbf{1.6}\times$), with Tensor Core (block-sparse) variants up to $\textbf{7.3}\times$ on locally dense graphs; for GATv2 we reach up to $\textbf{8.5}\times$ speedup (median $\textbf{2.0}\times$) while reducing peak memory by up to $\textbf{76}\times$ (median $\textbf{6}\times$). Our degree-aware reduction kernels achieve up to $\textbf{10}\times$ speedup (median $\textbf{2.6}\times$). For SpMM-based layers, properly cached cuSPARSE achieves up to $\textbf{8}\times$ speedup over DGL and outperforms evaluated custom baselines in the majority of evaluations. We release our implementations as drop-in replacements to support reproducible, hardware-aware GNN acceleration.

Related papers

• Scalable Graph Generative Modeling via Substructure Sequences [50.32639806800683]
  We introduce Generative Graph Pattern Machine (G$2$PM), a generative Transformer pre-training framework for graphs.<n>G$2$PM represents graph instances (nodes, edges, or entire graphs) as sequences of substructures.<n>It employs generative pre-training over the sequences to learn generalizable and transferable representations.
  arXiv  Detail & Related papers  (2025-05-22T02:16:34Z)
• GraNNite: Enabling High-Performance Execution of Graph Neural Networks on Resource-Constrained Neural Processing Units [0.6063137165121326]
  Graph Neural Networks (GNNs) are vital for learning from graph-structured data, enabling applications in network analysis, recommendation systems, and speech analytics.<n>GraNNite is the first hardware-aware framework optimizing GNN execution on commercial-off-the-shelf (COTS) SOTA accelerators.
  arXiv  Detail & Related papers  (2025-02-10T17:03:02Z)
• FastGL: A GPU-Efficient Framework for Accelerating Sampling-Based GNN Training at Large Scale [29.272368697268433]
  Graph Neural Networks (GNNs) have shown great superiority on non-Euclidean graph data. We propose FastGL, a GPU-efficient Framework for accelerating sampling-based training of GNN at Large scale. FastGL can achieve an average speedup of 11.8x, 2.2x and 1.5x over the state-of-the-art frameworks PyG, DGL, and GNNLab, respectively.
  arXiv  Detail & Related papers  (2024-09-23T11:45:47Z)
• Graph Transformers for Large Graphs [57.19338459218758]
  This work advances representation learning on single large-scale graphs with a focus on identifying model characteristics and critical design constraints. A key innovation of this work lies in the creation of a fast neighborhood sampling technique coupled with a local attention mechanism. We report a 3x speedup and 16.8% performance gain on ogbn-products and snap-patents, while we also scale LargeGT on ogbn-100M with a 5.9% performance improvement.
  arXiv  Detail & Related papers  (2023-12-18T11:19:23Z)
• Accel-GCN: High-Performance GPU Accelerator Design for Graph Convolution Networks [12.181052673940465]
  Graph Convolutional Networks (GCNs) are pivotal in extracting latent information from graph data across various domains. We present Accel-GCN, a GPU accelerator architecture for GCNs. Evaluation of Accel-GCN across 18 benchmark graphs reveals that it outperforms cuSPARSE, GNNAdvisor, and graph-BLAST by factors of 1.17 times, 1.86 times, and 2.94 times respectively.
  arXiv  Detail & Related papers  (2023-08-22T23:12:17Z)
• INR-Arch: A Dataflow Architecture and Compiler for Arbitrary-Order Gradient Computations in Implicit Neural Representation Processing [66.00729477511219]
  Given a function represented as a computation graph, traditional architectures face challenges in efficiently computing its nth-order gradient. We introduce INR-Arch, a framework that transforms the computation graph of an nth-order gradient into a hardware-optimized dataflow architecture. We present results that demonstrate 1.8-4.8x and 1.5-3.6x speedup compared to CPU and GPU baselines respectively.
  arXiv  Detail & Related papers  (2023-08-11T04:24:39Z)
• DistTGL: Distributed Memory-Based Temporal Graph Neural Network Training [18.52206409432894]
  DistTGL is an efficient and scalable solution to train memory-based TGNNs on distributed GPU clusters. In experiments, DistTGL achieves near-linear convergence speedup, outperforming state-of-the-art single-machine method by 14.5% in accuracy and 10.17x in training throughput.
  arXiv  Detail & Related papers  (2023-07-14T22:52:27Z)
• Communication-Efficient Graph Neural Networks with Probabilistic Neighborhood Expansion Analysis and Caching [59.8522166385372]
  Training and inference with graph neural networks (GNNs) on massive graphs has been actively studied since the inception of GNNs. This paper is concerned with minibatch training and inference with GNNs that employ node-wise sampling in distributed settings. We present SALIENT++, which extends the prior state-of-the-art SALIENT system to work with partitioned feature data.
  arXiv  Detail & Related papers  (2023-05-04T21:04:01Z)
• A Generalization of ViT/MLP-Mixer to Graphs [32.86160915431453]
  We introduce a new class of GNNs called Graph ViT/MLP-Mixer. They capture long-range dependency and mitigate the issue of over-squashing. They offer better speed and memory efficiency with a complexity linear to the number of nodes and edges.
  arXiv  Detail & Related papers  (2022-12-27T03:27:46Z)
• VersaGNN: a Versatile accelerator for Graph neural networks [81.1667080640009]
  We propose textitVersaGNN, an ultra-efficient, systolic-array-based versatile hardware accelerator. textitVersaGNN achieves on average 3712$times$ speedup with 1301.25$times$ energy reduction on CPU, and 35.4$times$ speedup with 17.66$times$ energy reduction on GPU.
  arXiv  Detail & Related papers  (2021-05-04T04:10:48Z)
• DistGNN: Scalable Distributed Training for Large-Scale Graph Neural Networks [58.48833325238537]
  Full-batch training on Graph Neural Networks (GNN) to learn the structure of large graphs is a critical problem that needs to scale to hundreds of compute nodes to be feasible. In this paper, we presentGNN that optimize the well-known Deep Graph Library (DGL) for full-batch training on CPU clusters. Our results on four common GNN benchmark datasets show up to 3.7x speed-up using a single CPU socket and up to 97x speed-up using 128 CPU sockets.
  arXiv  Detail & Related papers  (2021-04-14T08:46:35Z)
• Accelerated Gradient Tracking over Time-varying Graphs for Decentralized Optimization [59.65871549878937]
  We prove that the practical single loop accelerated gradient tracking needs $O(fracgamma1-sigma_gamma)2sqrtfracLepsilon)$.<n>Our convergence rates improve significantly over the ones of $O(frac1epsilon5/7)$ and $O(fracLmu)5/7frac1 (1-sigma)1.5logfrac1epsilon)$.
  arXiv  Detail & Related papers  (2021-04-06T15:34:14Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.