Everpure FlashBlade was purpose-built from the ground up to fully take advantage of flash for scale-out workloads and is optimized to support the data and GPU requirements for large-scale AI initiatives. FlashBlade allows capacity and performance to be scaled independently and can be continuously improved over time without disruption. This unique modular storage architecture offers flexible consumption choices while minimizing waste by using less power and space. The distributed storage architecture of FlashBlade matches the distributed scale-out compute infrastructure of DGX SuperPOD. FlashBlade is simple to manage; space and energy efficient; performant across different data types, IO profiles and AI workloads; and future-proof to help you meet changing AI requirements.