English
Related papers

Related papers: Breadth-First Pipeline Parallelism

200 papers

PipeDream is a Deep Neural Network(DNN) training system for GPUs that parallelizes computation by pipelining execution across multiple machines. Its pipeline parallel computing model avoids the slowdowns faced by data-parallel training when…

Distributed, Parallel, and Cluster Computing · Computer Science 2018-06-12 Aaron Harlap , Deepak Narayanan , Amar Phanishayee , Vivek Seshadri , Nikhil Devanur , Greg Ganger , Phil Gibbons

Many state-of-the-art ML results have been obtained by scaling up the number of parameters in existing models. However, parameters and activations for such large models often do not fit in the memory of a single accelerator device; this…

Machine Learning · Computer Science 2021-07-23 Deepak Narayanan , Amar Phanishayee , Kaiyu Shi , Xie Chen , Matei Zaharia

Pipeline parallelism is an essential technique in the training of large-scale Transformer models. However, it suffers from imbalanced memory consumption, leading to insufficient memory utilization. The BPipe technique was proposed to…

Machine Learning · Computer Science 2024-01-05 Mincong Huang , Chao Wang , Chi Ma , Yineng Zhang , Peng Zhang , Lei Yu

With the rapid adoption of large language models (LLMs) in recommendation systems, the computational and communication bottlenecks caused by their massive parameter sizes and large data volumes have become increasingly prominent. This paper…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-06-25 Haowei Yang , Yu Tian , Zhongheng Yang , Zhao Wang , Chengrui Zhou , Dannier Li

Training Deep Neural Networks (DNNs) with billions of parameters generally involves pipeline-parallel (PP) execution. Unfortunately, PP model training can use GPUs inefficiently, especially at large scale, due to idle GPU time caused by…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-10-11 Daiyaan Arfeen , Zhen Zhang , Xinwei Fu , Gregory R. Ganger , Yida Wang

Pipeline parallelism is widely used to train large language models (LLMs). However, increasing heterogeneity in model architectures exacerbates pipeline bubbles, thereby reducing training efficiency. Existing approaches overlook the…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-09-30 Jihu Guo , Tenghui Ma , Wei Gao , Peng Sun , Jiaxing Li , Xun Chen , Yuyang Jin , Dahua Lin

Pipeline parallelism is essential for large-scale model training, but existing asynchronous approaches often degrade convergence due to parameter mismatch between forward and backward passes. We propose Asynchronous Multi-Directional…

Distributed, Parallel, and Cluster Computing · Computer Science 2026-05-29 Ling Chen , Houming Wu , Wenjie Yu

We propose XPipe, an efficient asynchronous pipeline model parallelism approach for multi-GPU DNN training. XPipe is designed to use multiple GPUs to concurrently and continuously train different parts of a DNN model. To improve GPU…

Machine Learning · Computer Science 2020-11-10 Lei Guan , Wotao Yin , Dongsheng Li , Xicheng Lu

Recent hardware developments have dramatically increased the scale of data parallelism available for neural network training. Among the simplest ways to harness next-generation hardware is to increase the batch size in standard mini-batch…

Machine Learning · Computer Science 2019-07-22 Christopher J. Shallue , Jaehoon Lee , Joseph Antognini , Jascha Sohl-Dickstein , Roy Frostig , George E. Dahl

In this paper we analyze, evaluate, and improve the performance of training generalized linear models on modern CPUs. We start with a state-of-the-art asynchronous parallel training algorithm, identify system-level performance bottlenecks,…

Machine Learning · Computer Science 2018-12-20 Nikolas Ioannou , Celestine Dünner , Kornilios Kourtis , Thomas Parnell

Pipeline parallelism has been demonstrated to be a remarkable approach to improve throughput for training deep neural networks with billions of parameters over heterogeneous clusters. The 1F1B scheduling plan is a widely adopted strategy…

Distributed, Parallel, and Cluster Computing · Computer Science 2023-03-06 Siyu Wang , Zongyan Cao , Chang Si , Lansong Diao , Jiamang Wang , Wei Lin

Pipeline parallelism enables efficient training of Large Language Models (LLMs) on large-scale distributed accelerator clusters. Yet, pipeline bubbles during startup and tear-down reduce the utilization of accelerators. Although efficient…

Machine Learning · Computer Science 2023-05-16 Kazuki Osawa , Shigang Li , Torsten Hoefler

Training large deep learning models at scale is very challenging. This paper proposes Chimera, a novel pipeline parallelism scheme which combines bidirectional pipelines for efficiently training large-scale models. Chimera is a synchronous…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-08-22 Shigang Li , Torsten Hoefler

Transformer models have achieved state-of-the-art performance on various domains of applications and gradually becomes the foundations of the advanced large deep learning (DL) models. However, how to train these models over multiple GPUs…

Machine Learning · Computer Science 2022-11-28 Xupeng Miao , Yujie Wang , Youhe Jiang , Chunan Shi , Xiaonan Nie , Hailin Zhang , Bin Cui

Deep learning models trained on large data sets have been widely successful in both vision and language domains. As state-of-the-art deep learning architectures have continued to grow in parameter count so have the compute budgets and times…

Optimizing the parallel training of large models requires exploring intra-operator parallelism plans for a computation graph that typically contains tens of thousands of primitive operators. While the optimization of parallel data…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-07-08 Weifang Hu , Xuanhua Shi , Yunkai Zhang , Chang Wu , Xuan Peng , Jiaqi Zhai , Hai Jin , Xuehai Qian , Jingling Xue , Yongluan Zhou

In large language model (LLM) training, several parallelization strategies, including Tensor Parallelism (TP), Pipeline Parallelism (PP), Data Parallelism (DP), as well as Sequence Parallelism (SP) and Context Parallelism (CP), are employed…

Machine Learning · Computer Science 2024-11-12 Kazuki Fujii , Kohei Watanabe , Rio Yokota

DNN learning jobs are common in today's clusters due to the advances in AI driven services such as machine translation and image recognition. The most critical phase of these jobs for model performance and learning cost is the tuning of…

Distributed, Parallel, and Cluster Computing · Computer Science 2020-10-05 Isabelly Rocha , Nathaniel Morris , Lydia Y. Chen , Pascal Felber , Robert Birke , Valerio Schiavoni

Asynchronous pipeline model parallelism with a "1F1B" (one forward, one backward) schedule generates little bubble overhead and always provides quite a high throughput. However, the "1F1B" schedule inevitably leads to weight inconsistency…

Machine Learning · Computer Science 2025-02-18 Lei Guan , Dongsheng Li , Yongle Chen , Jiye Liang , Wenjian Wang , Xicheng Lu

Deploying deep learning (DL) models across multiple compute devices to train large and complex models continues to grow in importance because of the demand for faster and more frequent training. Data parallelism (DP) is the most widely used…

Machine Learning · Computer Science 2022-11-08 Saptadeep Pal , Eiman Ebrahimi , Arslan Zulfiqar , Yaosheng Fu , Victor Zhang , Szymon Migacz , David Nellans , Puneet Gupta