AI requires fast block-based storage for fast read and write access to and from the GPU’s. real time analytics demand ultra-low latency and fast reads to get results instantly.