华瑞指数云发布AI原生数据平台WADP：采用统一通用存储架构，实现AI算力与数据存力协同

Core Insights - ExponTech has launched a new unified AI data platform called WADP, aimed at addressing the core challenges of enterprise AI applications by efficiently integrating and managing vast amounts of multi-source data [1][2] - The platform is designed to facilitate seamless collaboration between AI computing power and data storage, ushering in a new era of "data on demand" for AI applications [1] Group 1: Key Features of WADP - WADP is built on a fully self-developed distributed storage engine and distributed KV metadata engine, capable of running on standard server hardware and adaptable to various general-purpose hardware [2] - The platform integrates core enterprise production systems with AI data pipelines, modernizing and replacing traditional storage arrays, file systems, and big data storage [2] - ExponTech emphasizes that its core code is entirely self-developed, avoiding reliance on foreign open-source technologies [2] Group 2: Market Trends and Needs - The demand for data storage is evolving due to the large-scale application of AI models, which is fundamentally reshaping the core requirements for data storage [1] - Four key trends are identified: performance enhancement to meet AI scenarios' high bandwidth and low latency needs, data integration for unified storage and management of heterogeneous data, intelligent collaboration to improve inference efficiency, and security control to ensure data safety [1] - The CEO of ExponTech highlights the necessity for deep collaboration between computing power and storage to eliminate data silos and enable seamless integration of core data with AI capabilities [2]