When Scaling Fails: Network and Fabric Effects on Distributed GPU Training Performance

• URL: http://arxiv.org/abs/2603.04424v1
• Date: Mon, 16 Feb 2026 16:19:47 GMT
• Title: When Scaling Fails: Network and Fabric Effects on Distributed GPU Training Performance
• Abstract summary: This paper examines why scaling fails in real systems, with a focus on the role of network and fabric effects.<n>We present an empirical study of distributed GPU training performance across multiple production-scale clusters.<n>We identify recurring failure modes that emerge as training transitions from single-node to multi-node execution.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Scaling distributed GPU training is commonly assumed to yield predictable performance gains as additional nodes are added. In practice, many large-scale deployments encounter diminishing returns and unstable behavior well before theoretical limits are reached. This paper examines why scaling fails in real systems, with a focus on the role of network and fabric effects that are often overlooked by higher-level training frameworks. We present an empirical study of distributed GPU training performance across multiple production-scale clusters. Our results show that network topology, congestion dynamics, collective synchronization behavior, and GPU locality frequently dominate end-to-end training performance once workloads move beyond a small number of nodes. Identical models and software stacks can exhibit sharply different scaling characteristics depending on fabric design and runtime communication patterns. We identify recurring failure modes that emerge as training transitions from single-node to multi-node execution, including synchronization amplification, topology-induced contention, and locality-driven performance variance. These effects are often invisible to standard profiling tools and are therefore misdiagnosed as framework or model-level inefficiencies. Based on these findings, we outline practical diagnostic principles that system builders can apply to understand scaling limits, improve predictability, and reduce the cost of large-scale distributed training.

Related papers

• Characterizing the Efficiency of Distributed Training: A Power, Performance, and Thermal Perspective [6.51239603014107]
  Large Language Models (LLMs) have pushed training workloads beyond the limits of single-node analysis.<n>We present a comprehensive characterization of LLM training across diverse real-world workloads and hardware platforms.
  arXiv  Detail & Related papers  (2025-09-12T16:05:07Z)
• Simplifying Distributed Neural Network Training on Massive Graphs: Randomized Partitions Improve Model Aggregation [23.018715954992352]
  We present a simplified framework for distributed GNN training that does not rely on the aforementioned costly operations. Specifically, our framework assembles independent trainers, each of which asynchronously learns a local model on locally-available parts of the training graph. In experiments on social and e-commerce networks with up to 1.3 billion edges, our proposed RandomTMA and SuperTMA approaches achieve state-of-the-art performance and 2.31x speedup compared to the fastest baseline.
  arXiv  Detail & Related papers  (2023-05-17T01:49:44Z)
• Decouple Graph Neural Networks: Train Multiple Simple GNNs Simultaneously Instead of One [60.5818387068983]
  Graph neural networks (GNN) suffer from severe inefficiency. We propose to decouple a multi-layer GNN as multiple simple modules for more efficient training. We show that the proposed framework is highly efficient with reasonable performance.
  arXiv  Detail & Related papers  (2023-04-20T07:21:32Z)
• Analyzing the Performance of Deep Encoder-Decoder Networks as Surrogates for a Diffusion Equation [0.0]
  We study the use of encoder-decoder convolutional neural network (CNN) as surrogates for steady-state diffusion solvers. Our results indicate that increasing the size of the training set has a substantial effect on reducing performance fluctuations and overall error.
  arXiv  Detail & Related papers  (2023-02-07T22:53:19Z)
• Intelligence Processing Units Accelerate Neuromorphic Learning [52.952192990802345]
  Spiking neural networks (SNNs) have achieved orders of magnitude improvement in terms of energy consumption and latency. We present an IPU-optimized release of our custom SNN Python package, snnTorch.
  arXiv  Detail & Related papers  (2022-11-19T15:44:08Z)
• Distributed Adversarial Training to Robustify Deep Neural Networks at Scale [100.19539096465101]
  Current deep neural networks (DNNs) are vulnerable to adversarial attacks, where adversarial perturbations to the inputs can change or manipulate classification. To defend against such attacks, an effective approach, known as adversarial training (AT), has been shown to mitigate robust training. We propose a large-batch adversarial training framework implemented over multiple machines.
  arXiv  Detail & Related papers  (2022-06-13T15:39:43Z)
• Distribution Mismatch Correction for Improved Robustness in Deep Neural Networks [86.42889611784855]
  normalization methods increase the vulnerability with respect to noise and input corruptions. We propose an unsupervised non-parametric distribution correction method that adapts the activation distribution of each layer. In our experiments, we empirically show that the proposed method effectively reduces the impact of intense image corruptions.
  arXiv  Detail & Related papers  (2021-10-05T11:36:25Z)
• Using Graph Neural Networks to model the performance of Deep Neural Networks [2.1151356984322307]
  We develop a novel performance model that adopts a graph representation. Experimental evaluation shows a 7:75x and 12x reduction in prediction error compared to the Halide and TVM models, respectively.
  arXiv  Detail & Related papers  (2021-08-27T20:20:17Z)
• Continuous-in-Depth Neural Networks [107.47887213490134]
  We first show that ResNets fail to be meaningful dynamical in this richer sense. We then demonstrate that neural network models can learn to represent continuous dynamical systems. We introduce ContinuousNet as a continuous-in-depth generalization of ResNet architectures.
  arXiv  Detail & Related papers  (2020-08-05T22:54:09Z)
• DBS: Dynamic Batch Size For Distributed Deep Neural Network Training [19.766163856388694]
  We propose the Dynamic Batch Size (DBS) strategy for the distributedtraining of Deep Neural Networks (DNNs) Specifically, the performance of each worker is evaluatedfirst based on the fact in the previous epoch, and then the batch size and dataset partition are dynamically adjusted. The experimental results indicate that the proposed strategy can fully utilizethe performance of the cluster, reduce the training time, and have good robustness with disturbance by irrelevant tasks.
  arXiv  Detail & Related papers  (2020-07-23T07:31:55Z)
• Understanding the Effects of Data Parallelism and Sparsity on Neural Network Training [126.49572353148262]
  We study two factors in neural network training: data parallelism and sparsity. Despite their promising benefits, understanding of their effects on neural network training remains elusive.
  arXiv  Detail & Related papers  (2020-03-25T10:49:22Z)

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.