IBM continues to progress with its mission of making artificial intelligence integral to storage. The company has been aggressively moving in this direction over the past few years, building AI into many of its solutions.
As part of its storage AI vision, IBM this week announced both a new addition to its Elastic Storage System product line and a reworked Cloud Object Storage (COS) storage engine. The IBM Elastic Storage System 5000, designed for data lakes with 55GB/s performance, joins the 3000 model, introduced in October. Both are powered by IBM Spectrum Scale software. The all-hard drive ESS 5000 can scale to yottabyte capacity in a single name space and incorporate object storage into its global namespace. It also can be integrated with the ESS 3000.
The software portion of the IBM COS storage engine has been upgraded, increasing system performance to 55GB/s in a multi-node cluster configuration. IBM says this can improve reads by 300% and writes by 15%, depending on object size. The company also has added a large-capacity hard drive for organizations that opt for IBM Cloud Object Storage as an array configuration. IBM COS also can now be incorporated into the Spectrum Scale global name space.
Finally, IBM has updated its Spectrum Discover metadata cataloging and indexing software. The software can now be deployed in a Red Hat OpenShift container environment as well as virtual machines.
While these announcements are not revolutionary, they are evolutionary, and they demonstrate the importance of AI and storage, said Scott Sinclair, a senior storage analyst at Enterprise Strategy Group.
AI and storage are a natural fit, he said. By optimizing storage to work well with AI, businesses can extract more value out of their data. And as data grows exponentially, companies are rapidly turning to AI to gain insight into their data. "Their strategy of innovating on multiple lines, including both data collection and analysis, and continuing to adjust along the way makes a lot of sense," he said.
IBM's plan is to continue down the storage AI path, said Eric Herzog, the company's storage chief marketing officer. While Herzog didn't give specific details of IBM's AI storage roadmap, he did say that IBM Storage will continue to deliver value for AI, big data and analytic applications and workloads. "We truly believe every company will become an AI company if they want to scale in the future," he said.