Morphling: Fast, Fused, and Flexible GNN Training at Scale

• URL: http://arxiv.org/abs/2512.01678v3
• Date: Fri, 05 Dec 2025 16:07:38 GMT
• Title: Morphling: Fast, Fused, and Flexible GNN Training at Scale
• Authors: Anubhab, Rupesh Nasre,
• Abstract summary: We present Morphling, a domain-specific code synthesizer designed to bridge this gap.<n>Morphling compiles high-level GNN into portable, specialized backends targeting OpenMP, MPI and MPI MPI.<n>It improves per-epoch training throughput by an average of 20X on CPU, 19X on GPU, and 6X in distributed settings over PyG and DGL, with peak speedups reaching 66X.
• Score: 0.3437656066916039
• License: http://creativecommons.org/publicdomain/zero/1.0/
• Abstract: Graph Neural Networks (GNNs) present a fundamental hardware challenge by fusing irregular, memory-bound graph traversals with regular, compute-intensive dense matrix operations. While frameworks such as PyTorch Geometric (PyG) and Deep Graph Library (DGL) prioritize high-level usability, they fail to address these divergent execution characteristics. As a result, they rely on generic kernels that suffer from poor cache locality, excessive memory movement, and substantial intermediate allocations. To address these limitations, we present Morphling, a domain-specific code synthesizer designed to bridge this gap. Morphling compiles high-level GNN specifications into portable, backend-specialized implementations targeting OpenMP, CUDA, and MPI. It achieves this by instantiating a library of optimized, architecture-aware primitives tailored to each execution environment. Morphling also incorporates a runtime sparsity-aware execution engine that dynamically selects dense or sparse execution paths using input feature statistics, reducing unnecessary computation on zero-valued entries. We evaluate Morphling on eleven real-world datasets spanning diverse graph structures, feature dimensionalities, and sparsity regimes. Morphling improves per-epoch training throughput by an average of 20X on CPUs, 19X on GPUs, and 6X in distributed settings over PyG and DGL, with peak speedups reaching 66X. Morphling's memory-efficient layouts further reduce peak memory consumption by up to 15X, enabling large-scale GNN training on commodity hardware. These findings demonstrate that specialized, architecture-aware code synthesis provides an effective and scalable path toward high-performance GNN execution across diverse parallel and distributed platforms.

Related papers

• FlashRNN: I/O-Aware Optimization of Traditional RNNs on modern hardware [6.749483762719583]
  State-tracking capabilities are important for time-series tasks and logical reasoning.<n>Traditional RNNs like LSTMs and GRUs do have these capabilities at the cost of strictly sequential processing.<n>We show how fast these networks can get with our hardware-optimization FlashRNN in Triton and optimized kernels to the register level.
  arXiv  Detail & Related papers  (2024-12-10T18:50:37Z)
• MaxK-GNN: Extremely Fast GPU Kernel Design for Accelerating Graph Neural Networks Training [7.193336207798203]
  We present MaxK-GNN, an advanced high-performance GPU training system integrating algorithm and system innovation. Experiments show that MaxK-GNN system could approach the theoretical speedup limit according to Amdahl's law. We achieve comparable accuracy to SOTA GNNs, but at a significantly increased speed: 3.22/4.24 times speedup (vs. theoretical limits, 5.52/7.27 times) on Reddit.
  arXiv  Detail & Related papers  (2023-12-14T05:00:49Z)
• T-GAE: Transferable Graph Autoencoder for Network Alignment [79.89704126746204]
  T-GAE is a graph autoencoder framework that leverages transferability and stability of GNNs to achieve efficient network alignment without retraining. Our experiments demonstrate that T-GAE outperforms the state-of-the-art optimization method and the best GNN approach by up to 38.7% and 50.8%, respectively.
  arXiv  Detail & Related papers  (2023-10-05T02:58:29Z)
• 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)
• Instant Neural Graphics Primitives with a Multiresolution Hash Encoding [67.33850633281803]
  We present a versatile new input encoding that permits the use of a smaller network without sacrificing quality. A small neural network is augmented by a multiresolution hash table of trainable feature vectors whose values are optimized through a gradient descent. We achieve a combined speed of several orders of magnitude, enabling training of high-quality neural graphics primitives in a matter of seconds.
  arXiv  Detail & Related papers  (2022-01-16T07:22:47Z)
• 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)
• Efficient and Generic 1D Dilated Convolution Layer for Deep Learning [52.899995651639436]
  We introduce our efficient implementation of a generic 1D convolution layer covering a wide range of parameters. It is optimized for x86 CPU architectures, in particular, for architectures containing Intel AVX-512 and AVX-512 BFloat16 instructions. We demonstrate the performance of our optimized 1D convolution layer by utilizing it in the end-to-end neural network training with real genomics datasets.
  arXiv  Detail & Related papers  (2021-04-16T09:54:30Z)
• 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)
• BlockGNN: Towards Efficient GNN Acceleration Using Block-Circulant Weight Matrices [9.406007544032848]
  Graph Neural Networks (GNNs) are state-of-the-art algorithms for analyzing non-euclidean graph data. How to inference GNNs in real time has become a challenging problem for some resource-limited edge-computing platforms. We propose BlockGNN, a software- hardware co-design approach to realize efficient GNN acceleration.
  arXiv  Detail & Related papers  (2021-04-13T14:09:22Z)
• Adaptive Filters and Aggregator Fusion for Efficient Graph Convolutions [11.769185588579488]
  We present state-of-the-art performance with lower memory consumption and latency, along with characteristics suited to accelerator implementation. Our proposal uses memory proportional to the number of vertices in the graph, in contrast to competing methods which require memory proportional to the number of edges. We propose aggregator fusion, a technique to enable GNNs to significantly boost their representational power, with only a small increase in latency of 19% over standard sparse matrix multiplication.
  arXiv  Detail & Related papers  (2021-04-03T20:54:36Z)

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.