# Optimizing Network Performance for Distributed DNN Training on GPU Clusters: ImageNet/AlexNet Training in 1.5 Minutes

arXiv: https://arxiv.org/abs/1902.06855 · HTML: https://mlsanity.com/paper/1902.06855/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Feb 2019.

## Built on by (most cited first, 13 of 13)

- [FFCV: Accelerating Training by Removing Data Bottlenecks](https://mlsanity.com/paper/2306.12517/) (2023)
- [Towards Crowdsourced Training of Large Neural Networks using Decentralized Mixture-of-Experts](https://mlsanity.com/paper/2002.04013/) (2020)
- [Enable Deep Learning on Mobile Devices: Methods, Systems, and Applications](https://mlsanity.com/paper/2204.11786/) (2022)
- [CASSINI: Network-Aware Job Scheduling in Machine Learning Clusters](https://mlsanity.com/paper/2308.00852/) (2023)
- [THC: Accelerating Distributed Deep Learning Using Tensor Homomorphic Compression](https://mlsanity.com/paper/2302.08545/) (2023)
- [Large-Scale Training System for 100-Million Classification at Alibaba](https://mlsanity.com/paper/2102.06025/) (2021)
- [Accelerating Distributed ML Training via Selective Synchronization](https://mlsanity.com/paper/2307.07950/) (2023)
- [ILASR: Privacy-Preserving Incremental Learning for Automatic Speech Recognition at Production Scale](https://mlsanity.com/paper/2207.09078/) (2022)
- [Accelerating Distributed Deep Learning using Lossless Homomorphic Compression](https://mlsanity.com/paper/2402.07529/) (2024)
- [Flexible Communication for Optimal Distributed Learning over Unpredictable Networks](https://mlsanity.com/paper/2312.02493/) (2023)
- [EmbRace: Accelerating Sparse Communication for Distributed Training of NLP Neural Networks](https://mlsanity.com/paper/2110.09132/) (2021)
- [ByteComp: Revisiting Gradient Compression in Distributed Training](https://mlsanity.com/paper/2205.14465/) (2022)
- [CD-SGD: Distributed Stochastic Gradient Descent with Compression and Delay Compensation](https://mlsanity.com/paper/2106.10796/) (2021)

## Abstract (arXiv)

It is important to scale out deep neural network (DNN) training for reducing model training time. The high communication overhead is one of the major performance bottlenecks for distributed DNN training across multiple GPUs. Our investigations have shown that popular open-source DNN systems could only achieve 2.5 speedup ratio on 64 GPUs connected by 56 Gbps network. To address this problem, we propose a communication backend named GradientFlow for distributed DNN training, and employ a set of network optimization techniques. First, we integrate ring-based allreduce, mixed-precision training, and computation/communication overlap into GradientFlow. Second, we propose lazy allreduce to improve network throughput by fusing multiple communication operations into a single one, and design coarse-grained sparse communication to reduce network traffic by only transmitting important gradient chunks. When training ImageNet/AlexNet on 512 GPUs, our approach achieves 410.2 speedup ratio and completes 95-epoch training in 1.5 minutes, which outperforms existing approaches.