Komprise, a provider of a data storage platform, is making available a tool that makes it simpler for IT teams to assess how data is being stored in flash memory on any platform or cloud service to help identify opportunities to reduce costs.
Company president and COO Krishna Subramanian said the Komprise Flash Stretch tool will enable IT teams to reduce costs by consuming flash memory more efficiently at a time when artificial intelligence (AI) applications have increased demand to the point where a general shortage is leading to price increases. In fact, Gartner estimates there will be a 130% increase in DRAM and solid-state drive (SSD) prices by the end of 2026.
Unfortunately, as much as 70% of the data being stored in these systems is cold in the sense that it has not been accessed any time recently, said Subramanian. Moving that data would enable IT teams to reduce cost by making room for more hot data in a way that would eliminate the need to either acquire additional storage systems or consume additional cloud storage services, she added.
The Komprise Flash Stretch tool achieves that goal by identifying data by its type, age and usage, how much savings could be generated by freeing up additional capacity and the overhead that might be generated if a need arose to rehydrate data that had been moved to a secondary storage platform. It also identifies the ideal tiering policy by department or use case along with any savings that might be generated by backing that data up using the Komprise platform.
Unlike in previous eras where there has been a shortage of memory, the current issues are more structural. Expanding memory production requires years of capital-intensive development. At the same time, a shift toward more advanced forms of memory has added complexity. High-bandwidth memory (HBM), which is increasingly used in AI systems, is particularly challenging to produce due to its layered design and precision manufacturing requirements. Given the amount of demand for memory being generated by AI applications, more manufacturing capacity is being allocated to HBM rather than memory that is used more typically by enterprise IT systems.
Less clear, however, is the degree to which memory shortages might finally force those same enterprise IT organizations to revisit their overall approach to data management. Historically, many IT teams have not focused much time and effort on optimizing usage of memory and storage with predictable results. The Komprise platform is optimized for storing unstructured data in a way that automatically creates an index that makes it simpler to retrieve that data when needed.
It will, of course, require a significant amount of missionary work to get enterprise IT teams to have a renewed appreciation for the art of data management and storage system optimization. However, as the cost of memory continues to climb, it’s now more a question of when and to what degree enterprise IT teams realize how much of the IT budget is being consumed by systems that in many cases are fundamentally inefficient.
