Shulou(Shulou.com)11/24 Report--

Thanks to CTOnews.com netizens for the delivery of clues about the past. CTOnews.com, July 14, Huawei recently released new commercial AI storage products called "OceanStor A310 Deep Learning data Lake Storage" and "FusionCube A3000 training / push Super". Officials said that "these two products can provide new momentum for AI basic model training, industry model training, subdivision scene model training and reasoning."

▲ image source Huawei CTOnews.com collates and summarizes that:

OceanStor A310 deep learning data lake storage, mainly for the basic / industry large model data lake scene, from data collection, preprocessing to model training, reasoning application of AI whole process massive data management.

The OceanStor A310 single-frame 5U supports the industry's highest 400GB/s bandwidth and the highest performance of 12 million IOPS, and can scale linearly to 4096 nodes, enabling multi-protocol lossless interworking, officials said. The global file system GFS realizes cross-regional intelligent data weaving, simplifies the process of data collection, realizes near data preprocessing through near-memory computing, reduces data movement, and improves preprocessing efficiency by 30%.

FusionCube A3000 training / push all-in-one machine is mainly aimed at industry large model training / reasoning scenarios. For 10 billion model applications, it integrates OceanStor A300 high performance storage nodes, training / push nodes, switching equipment, AI platform software and management operation and maintenance software to provide large model partners with a bag-carrying deployment experience, realizing one-stop "out-of-the-box" delivery and deployment within 2 hours.

Officials say the all-in-one supports two flexible business models, including Huawei's one-stop solution and an one-stop solution for third-party partners that open computing, web and AI platform software. The training / push nodes and storage nodes of the all-in-one machine can be expanded horizontally independently to match the model requirements of different sizes. At the same time, FusionCube A3000 implements the sharing of GPU among multiple model training reasoning tasks through high-performance containers, increasing the resource utilization from 40% to more than 70%.

Zhou Yuefeng, president of Huawei's data storage product line, said: "in the big model era, data storage has become the key infrastructure of the AI big model. Huawei data storage will provide a variety of solutions and products for the AI big model era, and work with partners to overcome these difficulties and promote AI capacity."

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.