Sandisk : AI token generation growth continues to scale in era of AI inference

This is evidenced as part of Sandisk's Tokenomics 2030 outlook that estimates global AI systems generate roughly 5 billion tokens per second today and could approach 317 billion per second by 2030, or approximately 10 quintillion tokens worldwide annually.

As AI moves toward longer context and more persistent, agentic workflows, the ability to store, stage, access, and reuse data becomes increasingly central to infrastructure productivity. Enterprise SSDs can provide a larger-capacity tier for selected KV-cache data, helping preserve previously computed information rather than repeatedly recreating it, acting like "token batteries" to help increase the value of fixed infrastructure investments.
Useful token output requires the full infrastructure stack, including storage, to work efficiently together. Sandisk is working to deliver the capacity, performance, and efficiency AI infrastructure requires into its NAND flash and enterprise SSDs as part of the infrastructure that turns data into intelligence.
Read more on the Tokenomics 2030 outlook via the webpage and blog post.

