Western Digital unveils AI Data Cycle Storage Framework

Western Digital has unveiled a new AI Data Cycle Storage Framework designed to help users enhance their AI infrastructure. The company aims to assist clients in maximising their AI investments, improving efficiency, and reducing the total cost of ownership (TCO) related to AI workflows.

The framework aims to address the growing demands of AI technologies, which require substantial data storage capacity and performance to manage immense volumes of data. As AI models continue to develop, data storage systems must be robust enough to handle large, sophisticated computations and the associated data loads. Western Digital’s new framework is expected to offer customers an organised approach to manage massive data growth effectively.

In conjunction with the new framework, Western Digital has introduced advanced enterprise-class SSDs and HDDs. These products are designed to support the key workloads within the AI Data Cycle. The six-stage AI Data Cycle framework outlines the optimal storage mix required for scalable AI workloads, detailing how Western Digital's SSD and HDD offerings can meet these requirements.

The newly introduced high-performance PCIe Gen5 SSD caters to AI training and inference. With its high capacity of 64TB, this SSD is expected to be a cornerstone for rapid AI data lake solutions. Additionally, Western Digital has showcased the world’s highest-capacity ePMR, UltraSMR 32TB HDD intended for cost-effective, large-scale data storage.

"Generative AI is viewed as the next transformational technology, and efficient data storage plays a critical role," said Ed Burns, Research Director at IDC. "Storage significantly influences the speed, efficiency, and accuracy of AI models, especially as larger and higher-quality datasets become more prevalent."

Rob Soderbery, Executive Vice President at Western Digital’s Flash Business Unit, added, "As AI technologies become embedded across virtually every industry sector, storage has become an increasingly important component of the AI technology stack. Our new AI Data Cycle framework equips customers to build infrastructure that enhances the performance, scalability, and deployment of AI applications, underscoring our commitment to delivering unparalleled value."

Western Digital's new suite of enterprise AI storage solutions has been designed to meet the intense compute and storage demands of modern AI workloads. The Ultrastar DC SN861 SSD features industry-leading random read performance, projected best-in-class power efficiency, and capacities up to 16TB. It aims to deliver up to three times the random read performance of its predecessor, with ultra-low latency and high responsiveness suited for large language model training and inferencing.

Furthermore, the Ultrastar DC SN655 enterprise-class SSD range has been expanded, now featuring U.3 SSDs with capacities up to 64TB. These updates are designed to provide enhanced performance and capacity for AI data preparation and accelerated data lakes.

In addition, Western Digital is now sampling the highest-capacity 32TB ePMR enterprise-class HDD. This Ultrastar DC HC690 high-capacity UltraSMR HDD is targeted at hyperscale cloud and enterprise data centres where massive data storage and low TCO are essential. Benefiting from proven designs, the new 32TB drive offers seamless qualification, integration, and rapid deployment while maintaining superior dependability and reliability.

Soderbery further emphasised the unique infrastructure and compute requirements of each stage of the AI Data Cycle. "We aim to help customers unlock the transformative capabilities of AI through our growing portfolio, long-term roadmap, and constant innovation."