When it comes to AI use cases and applications, data storage isn’t optional—it’s foundational to progress and value. Of course, storage has always mattered greatly for all IT initiatives, but the technology has advanced substantially past the days of spindles and block and file storage. In fact, it’s fair to say that AI has dramatically changed the rules of the game when it comes to storage.

There are several key reasons for this, including the extensive volumes of data both created by different AI versions—generative, predictive, causal—and used in the creation of AI training models at the heart of these initiatives. But that’s not all: data collection, data preparation, training, experimentation, and inferencing/production all are pushing the limits when it comes to data volumes.

According to research by Informa TechTarget’s Enterprise Strategy Group, of those organizations that have an active or planned AI-centric multi-vendor evaluation project that requires data center storage and/or HCI, 67% cited data collection/preparation as the workload that is or will be supported by the initiative, followed by model development/training (64%), model deployment/inferencing (63%), and experimentation (46%).