MWC: Huawei Unveils Three Storage Solutions for AI Era

At MWC Barcelona 2024, Dr. Peter Zhou, president, IT product line, Huawei Technologies Co., Ltd. launched 3 solutions aimed at helping carriers WW build data infrastructure in the AI era: AI data lake solution, all-scenario data protection solution, and DCS full-stack data center solution.

Peter Zhou delivering keynote speech at MWC

The rapid development of large AI models is unleashing the value of data assets and paving the way for an era of data awakening. There are 2 major challenges in data infrastructure construction in the AI era.

First, to extract knowledge from data, independent and scattered data sources must be aggregated in AI data factories to mine value from the data. This sets high requirements for data mobility.

Second, data assetization is growing at a rapid pace. The average retention time of annotated AI data and model data is now more than 3 years, and some valuable corpus data needs to be permanently stored.

To address these challenges, the company has launched a series of storage products and solutions to help carriers build leading data infrastructure in the AI era.

• AI data lake solution for transferring mass data to valuable assets: This solution includes Data Management Engine (DME), a 1-stop knowledge generation tool chain (eDataMate), data fabric, and the AI storage solutions OceanStor A800 and OceanStor A310. It makes AI training data globally visible, manageable, and available, and improves data collection, preprocessing, and training efficiency. In particular, OceanStor A800, a next-gen high-performance NAS storage solution, is an AI storage power foundation with a data and control plane separation architecture, the OceanFS high-performance parallel file system, flexible bidirectional expansion, and other capabilities, all designed to support real-time hyper-cluster scheduling.

• All-scenario data protection solution for data production, backup, and archiving that boosts the resilience of data assets: This solution uses 2 layers of collaborative deployment. The 1st layer is production-backup collaboration, backing massive numbers of small files up 90% faster than conventional solutions. It also supports real-time native-format data analytics. Then is the backup-archive collaboration. Its post-de-dupe tiering feature reduces bandwidth consumption by 80%. Additionally, the company launched the OceanProtect E8000 and X9000 backup appliance solutions and next-gen OceanStor Arctic magneto-electric storage solution that is designed for warm and cold data at the event. The magneto-electric storage solution is predicted to reduce TCO by 20% than tapes and power consumption by 90% than HDDs.

• DCS full-stack data center solution for helping carriers achieve higher IT service efficiency and agility: This full-stack solution, built on virtualization and containers, covers data collection, storage, computing, management, and usage, delivering comprehensive IaaS and PaaS.

Zhou said that AI is prompting a data awakening, which is accelerating the data assetization process. Huawei Data Storage will remain committed to providing storage solutions for all scenarios, to help carriers build leading data infrastructure in the AI era.