Western Digital has unveiled a new AI Data Cycle framework aimed at optimising storage for AI workloads, enhancing efficiency, and reducing costs for businesses. Tailored to meet the demanding requirements of AI data processing and generation, this framework promises to significantly improve how AI models handle data, ensuring faster performance, better scalability, and lower total costs of ownership.
The six-stage AI Data Cycle framework aligns with Western Digital's growing portfolio of enterprise-class SSDs and HDDs, enabling businesses to harness the accelerated compute capacity needed to fuel and deploy AI applications. Key innovations presented include Western Digital's first high-performance PCIe Gen5 SSD, a 64TB SSD high-capacity solution for rapid AI data lakes, and the UltraSMR 32TB HDD, the world's highest capacity ePMR drive.
Ed Burns, Research Director at IDC, commented on the development, stating, "There's no doubt that generative AI is the next transformational technology, and storage is a critical enabler. The implications for storage are expected to be significant, as the role of storage and access to data influences the speed, efficiency, and accuracy of AI models, especially as larger and higher-quality data sets become more prevalent." He added that Western Digital is well-positioned to benefit in the growing AI landscape with its strong market position and extensive portfolio catering to various stages of the AI data cycle.
Rob Soderbery, Executive Vice President and General Manager of Western Digital's Flash Business Unit, echoed Burns' sentiments, stating, "Data is the fuel of AI. As AI technologies become embedded across virtually every industry sector, storage has become an increasingly important and dynamic component of the AI technology stack." He asserted that the new AI Data Cycle framework would equip clients to build a storage infrastructure that enhances performance, scalability, and deployment of AI applications, underscoring Western Digital's commitment to delivering unparalleled value.
The Ultrastar DC SN861 SSD, Western Digital's first enterprise-class PCIe Gen 5.0 solution, is one of the flagship innovations announced. It offers industry-leading random read performance and projected best-in-class power efficiency for AI workloads. With capacities up to 16TB, it promises up to a threefold increase in random read performance compared with previous generations. The SSD is designed for mission-critical workloads, featuring NVMe 2.0 and OCP 2.0 support, and is now in the sampling phase.
Complementing this, the Ultrastar DC SN655 enterprise-class SSD range, aimed at storage-intensive applications, has also been expanded. The new U.3 SSD options will reach up to 64TB, targeting higher performance and capacity for AI data preparation and larger data lakes. Similarly, the 32TB ePMR enterprise-class HDD, the Ultrastar DC HC690 high-capacity UltraSMR HDD, is now sampling and is designed for massive data storage in hyperscale cloud and enterprise data centres. This drive is crucial for AI workflows where large-scale storage and low TCO are paramount.
Soderbery further elaborated, "Each stage of the AI Data Cycle is unique with different infrastructure and compute requirements. By understanding the dynamic interplay between AI and data storage, Western Digital is delivering solutions that not only offer higher capacities but are also tailored to support the extreme performance and endurance of next-generation AI workloads." He emphasised the company's ongoing innovation and long-term roadmap aim to help customers unlock the transformative capabilities of AI.
Through these advancements, Western Digital aims to support key workloads within the AI Data Cycle and help customers capture the value of AI, fortifying its position in the competitive landscape of AI-driven data storage solutions. With a focus on reducing total costs of ownership and improving efficiency, Western Digital's new framework is set to be a pivotal development for businesses navigating the complexities of AI data processing and generation.
