IBM expands storage portfolio to drive AI deployments

IBM unveils new and updated storage products, including the new IBM Elastic Storage System 5000, which is scalable to yottabyte configurations.

IBM on the state of natural language processing

IBM on Thursday unveiled new and updated storage products designed to help enterprises build AI-optimized infrastructure. The expanded product lineup includes the new IBM Elastic Storage System 5000, which is scalable to yottabyte configurations.

In May, IBM CEO Arvind Krishna kicked off IBM's Think Digital 2020 conference by promising that IBM "will support your journeys" through hybrid cloud and AI and declaring that AI will be a key reason customers work with IBM. "Every company will become an AI company," he said. "Not because they can, but because they must."

18181354.jpg

The new Elastic Storage System 5000 (ESS 5000) is a software-defined storage solution designed for the first stage of an organization's "AI journey" -- data collection. Optimized for large capacity data lakes, it couples Spectrum Scale, IBM's high-performance clustered file system software, with IBM Power9 compute. 

ESS 5000 delivers 55GB/sec performance throughput per node and is scalable to yottabyte configurations. It can seamlessly integrate with a high-performance tier of storage, such as the ESS 3000, which IBM unveiled last year to meet the demands of data analysis.

Must read:

Additionally, IBM on Thursday announced it's updating IBM Cloud Object Storage (COS) with a new software engine for improved performance. The upgrade should improve reads by 300 percent and writes by 150 percent, depending on object size. The better performance could enable COS to support integration with AI and big data workflows. The updated COS will also support a new technology of high-capacity disk drives -- Shingled Magnetic Recording, or SMR drives -- providing 1.9 PB in a 4u disk enclosure. 

Lastly, IBM is deploying IBM Spectrum Discover, software designed to provide file and object data cataloging and indexing, on Red Hat OpenShift. This will help businesses make their data portable and easily deployed across clouds.