Getting to Know the Unsung Hero of AI Infrastructure with Solidigm

Solidigm’s presentation at AI Infrastructure Field Day emphasized the critical role of data storage in the evolving landscape of AI solutions, using the framework of the AI data pipeline. The amount of data used for training an AI model directly correlates with its accuracy, making storage a critical infrastructure component. Ace Stryker, the Director of Market Development at Solidigm, introduced the company’s focus on storage solutions, explaining the importance of efficient data storage in all stages of the AI data pipeline, from data ingestion and preparation to model development, inference, and archiving.

The presentation shows Solidigm as a key player in AI infrastructure by providing a diverse portfolio of SSD offerings that address the growing demands for high-capacity and efficient storage. Ace explained that Solidigm, spun off from Intel’s storage and memory group, has a strong heritage in storage technology, emphasizing the importance of storage solutions in enabling more GPU scaling. Solidigm also introduced a key collaboration with NVIDIA that will enable future generations of storage solutions, specifically in liquid cooling, which allows for future storage solutions.

The presentation explored the nuances of the AI data pipeline, highlighting how storage solutions can improve model accuracy and address crucial challenges within AI infrastructure. The presentation’s core message revolves around how every watt and square inch of data center space counts, especially when tackling advanced AI applications requiring substantial data. The presentation also included a live demo from Metrum AI, illustrating data-intensive inference possibilities and discussing high-density storage solutions.