What is Shared Accelerated Storage?

Shared Accelerated Storage is a new class of storage that leverages NVMe and flash technology to enable diskless, stateless servers to connect to shared storage via fast networks in a parallel, optimised architecture from compute to flash chips.

How Shared Accelerated Storage Can Drive My Business

Shared Accelerated Storage provides a shared pool of flash-accelerated, highly-available, and easily-scalable storage to dozens of clients – centrally managed and connected via high-bandwidth, low-latency networks. It thus eliminates the traditional disadvantages of both DAS (limited capacity, managed as a silo) and SAN/NAS (poor reliability, complex at scale). Organisations get fast storage, simplified management, and open-ended scale.

Another benefit: data center compute and storage can now be disaggregated without performance loss. With the advent of fast networking protocols like NVMe-oF, Shared Accelerated Storage speaks to compute via massively parallel networks at speeds that are comparable to DAS, but without its limitations. In addition, Shared Accelerated Storage removes storage I/O processing loads from CPUs, giving you back up to 30% of your compute power.

Pure’s Take on Shared Accelerated Storage

We think of Shared Accelerated Storage, a term coined by Gartner, as a wholly new storage architecture – one that has the potential to unite the formerly disparate worlds of shared storage (SAN/NAS) and direct-attached storage (DAS). This enables classic scale-up applications like databases and VMs to share the same data-centric architecture as modern scale-out applications like analytics, NoSQL databases, and container farms.

The potential benefits for the enterprise are huge; thus, we believe that standardization on this new architecture can and will be swift, and we’re delivering Shared Accelerated Storage products today.

More For You

Explore Pure Data-Centric Architecture

The Era of Fast Networks
is Here

Unpack Shared Accelerated Storage for AI

+