Research on optimization of storage system in intelligent computing center

The intelligent computing center uses distributed file storage for data preprocessing and model training, distributed object storage for the acquisition of raw data and model release, and distributed block storage to provide storage for the resource management platform. Meanwhile, high-performance d...

Full description

Saved in:
Bibliographic Details
Main Authors: CAO Yuanming, LEI Ming, LIU Qin, NIU Yingxia, WU Zhenyu, PAN Jie
Format: Article
Language:zho
Published: Beijing Xintong Media Co., Ltd 2025-07-01
Series:Dianxin kexue
Subjects:
Online Access:http://www.telecomsci.com/zh/article/doi/10.11959/j.issn.1000-0801.2025160/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849398144655163392
author CAO Yuanming
LEI Ming
LIU Qin
NIU Yingxia
WU Zhenyu
PAN Jie
author_facet CAO Yuanming
LEI Ming
LIU Qin
NIU Yingxia
WU Zhenyu
PAN Jie
author_sort CAO Yuanming
collection DOAJ
description The intelligent computing center uses distributed file storage for data preprocessing and model training, distributed object storage for the acquisition of raw data and model release, and distributed block storage to provide storage for the resource management platform. Meanwhile, high-performance distributed file storage is used to shorten the read and write time of checkpoint during the training process and improve the training efficiency of the cluster. The entire life cycle of large model training requires data copying and migration between storage systems with different storage protocols and different read-write performances, resulting in duplicate data storage. Additionally, data copying requires computing resources and network bandwidth. To address the above issues and provide a unified namespace for the intelligent computing clusters, a file and object converged storage and a hierarchical file storage scheme were proposed to solve the problem of data transfer between different storage protocols and enable automatic data flow between high-performance file storage (all SSD) and ordinary-performance file storage (SSD and HDD), providing a reference for the optimization of storage systems of ultra-large-scale intelligent computing clusters.
format Article
id doaj-art-8a89a2e225ad40b8b047f7e0fc3a0022
institution Kabale University
issn 1000-0801
language zho
publishDate 2025-07-01
publisher Beijing Xintong Media Co., Ltd
record_format Article
series Dianxin kexue
spelling doaj-art-8a89a2e225ad40b8b047f7e0fc3a00222025-08-20T03:38:43ZzhoBeijing Xintong Media Co., LtdDianxin kexue1000-08012025-07-0141164175120127889Research on optimization of storage system in intelligent computing centerCAO YuanmingLEI MingLIU QinNIU YingxiaWU ZhenyuPAN JieThe intelligent computing center uses distributed file storage for data preprocessing and model training, distributed object storage for the acquisition of raw data and model release, and distributed block storage to provide storage for the resource management platform. Meanwhile, high-performance distributed file storage is used to shorten the read and write time of checkpoint during the training process and improve the training efficiency of the cluster. The entire life cycle of large model training requires data copying and migration between storage systems with different storage protocols and different read-write performances, resulting in duplicate data storage. Additionally, data copying requires computing resources and network bandwidth. To address the above issues and provide a unified namespace for the intelligent computing clusters, a file and object converged storage and a hierarchical file storage scheme were proposed to solve the problem of data transfer between different storage protocols and enable automatic data flow between high-performance file storage (all SSD) and ordinary-performance file storage (SSD and HDD), providing a reference for the optimization of storage systems of ultra-large-scale intelligent computing clusters.http://www.telecomsci.com/zh/article/doi/10.11959/j.issn.1000-0801.2025160/converged storagehierarchical storagedistributed file storagedistributed object storage
spellingShingle CAO Yuanming
LEI Ming
LIU Qin
NIU Yingxia
WU Zhenyu
PAN Jie
Research on optimization of storage system in intelligent computing center
Dianxin kexue
converged storage
hierarchical storage
distributed file storage
distributed object storage
title Research on optimization of storage system in intelligent computing center
title_full Research on optimization of storage system in intelligent computing center
title_fullStr Research on optimization of storage system in intelligent computing center
title_full_unstemmed Research on optimization of storage system in intelligent computing center
title_short Research on optimization of storage system in intelligent computing center
title_sort research on optimization of storage system in intelligent computing center
topic converged storage
hierarchical storage
distributed file storage
distributed object storage
url http://www.telecomsci.com/zh/article/doi/10.11959/j.issn.1000-0801.2025160/
work_keys_str_mv AT caoyuanming researchonoptimizationofstoragesysteminintelligentcomputingcenter
AT leiming researchonoptimizationofstoragesysteminintelligentcomputingcenter
AT liuqin researchonoptimizationofstoragesysteminintelligentcomputingcenter
AT niuyingxia researchonoptimizationofstoragesysteminintelligentcomputingcenter
AT wuzhenyu researchonoptimizationofstoragesysteminintelligentcomputingcenter
AT panjie researchonoptimizationofstoragesysteminintelligentcomputingcenter