Saved in:
Bibliographic Details
Main Authors: Olama, Alireza, Lundell, Andreas, Hajj, Izzat El, Lilius, Johan, Björkqvist, Jerker
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.14628
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917149105192960
author Olama, Alireza
Lundell, Andreas
Hajj, Izzat El
Lilius, Johan
Björkqvist, Jerker
author_facet Olama, Alireza
Lundell, Andreas
Hajj, Izzat El
Lilius, Johan
Björkqvist, Jerker
contents Inter-node communication bandwidth increasingly constrains distributed training at scale on multi-node GPU clusters. While compact models are the ultimate deployment target, conventional pruning-aware distributed training systems typically fail to reduce communication overhead because unstructured sparsity cannot be efficiently exploited by highly optimized dense collective primitives. We present PruneX, a distributed data-parallel training system that co-designs pruning algorithms with cluster hierarchy to reduce inter-node bandwidth usage. PruneX introduces the Hierarchical Structured ADMM (H-SADMM) algorithm, which enforces node-level structured sparsity before inter-node synchronization, enabling dynamic buffer compaction that eliminates both zero-valued transmissions and indexing overhead. The system adopts a leader-follower execution model with separated intra-node and inter-node process groups, performing dense collectives on compacted tensors over bandwidth-limited links while confining full synchronization to high-bandwidth intra-node interconnects. Evaluation on ResNet architectures across 64 GPUs demonstrates that PruneX reduces inter-node communication volume by approximately 60% and achieves 6.75x strong scaling speedup, outperforming the dense baseline (5.81x) and Top-K gradient compression (3.71x) on the Puhti supercomputer at CSC - IT Center for Science (Finland).
format Preprint
id arxiv_https___arxiv_org_abs_2512_14628
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PruneX: A Hierarchical Communication-Efficient System for Distributed CNN Training with Structured Pruning
Olama, Alireza
Lundell, Andreas
Hajj, Izzat El
Lilius, Johan
Björkqvist, Jerker
Distributed, Parallel, and Cluster Computing
Inter-node communication bandwidth increasingly constrains distributed training at scale on multi-node GPU clusters. While compact models are the ultimate deployment target, conventional pruning-aware distributed training systems typically fail to reduce communication overhead because unstructured sparsity cannot be efficiently exploited by highly optimized dense collective primitives. We present PruneX, a distributed data-parallel training system that co-designs pruning algorithms with cluster hierarchy to reduce inter-node bandwidth usage. PruneX introduces the Hierarchical Structured ADMM (H-SADMM) algorithm, which enforces node-level structured sparsity before inter-node synchronization, enabling dynamic buffer compaction that eliminates both zero-valued transmissions and indexing overhead. The system adopts a leader-follower execution model with separated intra-node and inter-node process groups, performing dense collectives on compacted tensors over bandwidth-limited links while confining full synchronization to high-bandwidth intra-node interconnects. Evaluation on ResNet architectures across 64 GPUs demonstrates that PruneX reduces inter-node communication volume by approximately 60% and achieves 6.75x strong scaling speedup, outperforming the dense baseline (5.81x) and Top-K gradient compression (3.71x) on the Puhti supercomputer at CSC - IT Center for Science (Finland).
title PruneX: A Hierarchical Communication-Efficient System for Distributed CNN Training with Structured Pruning
topic Distributed, Parallel, and Cluster Computing
url https://arxiv.org/abs/2512.14628