Pure Storage Announces FlashBlade//EXA: The Future of AI and HPC Scalability and Performance

Legacy storage systems were not designed to meet the modern requirements of AI. When applied to large-scale AI and HPC, they encounter critical limitations in parallel and concurrent reads and writes, metadata performance, ultra-low latency, asynchronous checkpointing, and predictable high throughput.

Customers wanting flexibility will benefit from a disaggregated architecture that maintains Pure Storage’s famous dependability, scalability, simplicity, and performance. This cutting-edge technology enables companies to swiftly adapt to new requirements while staying true to their core software and values. Built on the same principles and technology as Pure Storage hyperscaler design, DirectFlash and Purity now integrate in new ways, allowing Pure Storage to deliver the world’s highest performance storage system for large-scale AI/ML and high-performance computing.

Here are some of the exceptional features Pure Storage is delivering with FlashBlade//EXA:

Performance Density

FlashBlade//EXA will offer industry-leading performance in a compact form, delivering 3.4 TB/s per rack at GA this summer.

Unmatched Scale

FlashBlade//EXA will provide unparalleled namespace and capacity scaling, supporting trillions of files and exabytes of capacity at GA this summer.

Improved Adaptability

FlashBlade//EXA will give customers the flexibility to choose their own qualified data nodes, allowing for seamless and predictable scaling of read/write performance, metadata IOPS, and capacity.

Multi-Dimensional Performance

FlashBlade//EXA will deliver the extreme metadata performance required for multi-modal data access.

More Pure Storage products >>