IBM today refreshed its flash storage system portfolio to add a FlashCore Module all-flash solid disk drive (SSD) that provides access to 105TB of capacity, enabling up to 2.5PB of storage on a 1u system and up to 11.8PB in a 2u system.
At the same time, IBM is adding an artificial intelligence (AI) agent, dubbed FlashSystem.ai, that is designed to help storage administrators manage, monitor, diagnose, and remediate issues. The FlashSystem.ai is also designed to learn over time to adapt to application behavior. That capability makes it possible to suggest performance improvements based on reasoning the AI agent is able to explain.
Sam Werner, general manager of IBM Storage, said the IBM FlashSystem 5600, 7600, and 9600 systems are now the densest SSD arrays that are designed from the ground up for enterprise-class applications.
Additionally, IBM is also committed to remain a price leader at a time when the cost of memory has risen sharply, added Werner. Because IBM manufactures its own drives, IBM is in a better position to control costs, he added.
From a total cost perspective, IBM is also leading given the density of the arrays being added to its portfolio, noted Werner. In effect, IBM is now providing up to 40% greater data efficiency in a way that compared to the previous generation of IBM SSD system reduces the required storage footprint by 30%-75%, via optimized placement of data and platform consolidation, he noted.
Achieving that level of optimization will, of course, become much easier with the help of AI agents. IBM intends to make the AI agents it has developed to optimize storage systems across its entire portfolio, said Werner. The overall goal is to make it simpler for storage administrators to optimally store data in ways that both increase performance while reducing total costs, he added.
Of course, it won’t be too long before every storage system provides access to some type of AI agent that reduces the level of expertise required to optimally manage data. In time, those AI agents will also communicate with AI agents that have been deployed to not just manage servers but also the software deployed on them. As a result, IT teams should be able to define a set of goals and metrics they want to achieve and maintain, which will then be automatically implemented by multiple AI agents that communicate directly with one another.
Less clear is the impact those AI agents will have on how internal IT teams are structured today. As it becomes possible to better unify the management of IT infrastructure and applications, the way IT teams have been constructed around various specialized skill sets is likely to change.
In the meantime, however, it’s clear that the amount of data that IT teams are expected to manage is only going to continue to exponentially increase in the age of AI. In fact, it’s arguable that without some help from AI agents it will not be possible for mere mortal storage administrators to manage what is rapidly becoming an overwhelming amount of data.
