Usually, enterprises generate large amounts of data but one of the significant problems in this is that the data that is being stored must also be processed. Storage architectures are known for having storage capacity problems when the flash storage was introduced, storage received a much-needed boost in performance and this resulted in an increase in transactional workloads like databases as well as block, object, and file storage. In recent times workloads are starting to incorporate new techniques like AI and advanced analytics. When the data is moved we must consider two factors with are cost and time.
A small group of startup companies has pioneered something in the emerging field called computational storage or in-situ processing. This group has added several new kinds of approaches to computational storage. New approaches revolve mostly around NV-Express SSDs and they add accelerator, integrate processing power into individual drives and so on. The Storage Networking Industry Association (SNIA) formed a new group even though the technology is still developing. The group is called Computational Storage Technical Work Group (TWG) and it was formed in 2018 to promote interoperability. SNIA stated that the startup in the group focuses on working on one-off proof-of-concept deployments that might eventually gain traction among investors.