1. Fire-Flyer AI-HPC: A Cost-Effective Software-Hardware Co-Design for Deep Learning
- Author
-
An, Wei, Bi, Xiao, Chen, Guanting, Chen, Shanhuang, Deng, Chengqi, Ding, Honghui, Dong, Kai, Du, Qiushi, Gao, Wenjun, Guan, Kang, Guo, Jianzhong, Guo, Yongqiang, Fu, Zhe, He, Ying, Huang, Panpan, Li, Jiashi, Liang, Wenfeng, Liu, Xiaodong, Liu, Xin, Liu, Yiyuan, Liu, Yuxuan, Lu, Shanghao, Lu, Xuan, Nie, Xiaotao, Pei, Tian, Qiu, Junjie, Qu, Hui, Ren, Zehui, Sha, Zhangli, Su, Xuecheng, Sun, Xiaowen, Tan, Yixuan, Tang, Minghui, Wang, Shiyu, Wang, Yaohui, Wang, Yongji, Xie, Ziwei, Xiong, Yiliang, Xu, Yanhong, Ye, Shengfeng, Yu, Shuiping, Zha, Yukun, Zhang, Liyue, Zhang, Haowei, Zhang, Mingchuan, Zhang, Wentao, Zhang, Yichao, Zhao, Chenggang, Zhao, Yao, Zhou, Shangyan, Zhou, Shunfeng, and Zou, Yuheng more...
- Subjects
Computer Science - Distributed, Parallel, and Cluster Computing ,Computer Science - Artificial Intelligence - Abstract
The rapid progress in Deep Learning (DL) and Large Language Models (LLMs) has exponentially increased demands of computational power and bandwidth. This, combined with the high costs of faster computing chips and interconnects, has significantly inflated High Performance Computing (HPC) construction costs. To address these challenges, we introduce the Fire-Flyer AI-HPC architecture, a synergistic hardware-software co-design framework and its best practices. For DL training, we deployed the Fire-Flyer 2 with 10,000 PCIe A100 GPUs, achieved performance approximating the DGX-A100 while reducing costs by half and energy consumption by 40%. We specifically engineered HFReduce to accelerate allreduce communication and implemented numerous measures to keep our Computation-Storage Integrated Network congestion-free. Through our software stack, including HaiScale, 3FS, and HAI-Platform, we achieved substantial scalability by overlapping computation and communication. Our system-oriented experience from DL training provides valuable insights to drive future advancements in AI-HPC., Comment: This is the preprint version of the paper accepted for presentation at the 2024 International Conference for High Performance Computing, Networking, Storage, and Analysis (SC'24). \c{opyright} 2024 IEEE. Personal use of this material is permitted. For other uses, permission from IEEE must be obtained. Please refer to IEEE Xplore for the final published version more...
- Published
- 2024