AI Insights
Showing results for 
Search instead for 
Did you mean: 

Persistent Memory: The Key Driver Behind IT Agility


Database storage bottlenecks have hampered business users in getting faster access to their data. HPE Scalable Persistent Memory eliminates storage bottlenecks by running at memory speeds with the persistence of storage.

Blog_PersistentMemory.jpgBusiness users are getting frustrated by delays in getting data from their business applications caused by database bottlenecks. When they can’t get the information they need to in real time to run analytics and to do their jobs, sales cycles slow down, orders are delayed, and customer service suffers.

Many of these bottlenecks are the result of traditional compute paradigms. Server memory chips move data fast but can only retain the data as long as the application is open and the power is on. Conversely, HDDs and SSDs are much slower than memory, but they retain the data after applications are closed and the server is turned off.

Servers need to move the data back-and-forth between memory and storage media as business users interact with the files. As volumes increase rapidly—thanks to trends like cloud solutions, 24x7 mobility, and Big Data analytics—bottlenecks ensue. When information can’t move fast enough, business slows down.

Persistent memory to the rescue

Persistent memory is resolving this issue for many businesses. This compute paradigm creates greater IT agility by combining the speed of memory with the persistence of storage to give businesses the best of both worlds: An IT infrastructure that can now keep pace with business demands by delivering low-latency application performance and built-in persistence—while still ensuring reliability and efficiency.

Persistent memory achieves this impressive feat with dynamic random access memory (DRAM). A portion of the DRAM is designated as persistent memory for application acceleration with a dedicated flash storage tier for data backup and a backup power source to facilitate moving data from DRAM to flash. This enables a terabyte-scale storage solution running at DRAM speeds eliminating storage bottlenecks that typically result from writing to slower storage media.

Checkpoint operations occur whenever an application that has data in-flight in memory periodically saves to non-volatile storage.  This is a very common storage bottleneck resulting in spikes in application performance.  HPE Persistent Memory dramatically improves and smooths out these operations dramatically improving overall application performance.. This ensures optimal application performance and real-time analytics, regardless of the application payload.

The persistent memory paradigm also accelerates restores. Businesses that need to recover data following a server crash or hard drive failure can recover more quickly and reestablish normal business operations right away – at the speed of memory.

Industry collaboration ensures fast and reliable application performance

HPE is pioneering the IT transformation to persistent memory. HPE Scalable Persistent Memory inside HPE ProLiant Gen10 servers runs at memory speeds while providing terabyte scale capacity. This enables businesses to unlock new levels of application performance  and workload optimization:

  • 4x faster database transaction log writes compared to solid-state drives
  • 27x faster checkpoint operations
  • 20x faster database restarts vs. flash media alone

In developing persistent memory, HPE has worked side by side with the leading operating system and enterprise application providers to create a software ecosystem that synchronizes with the technology. This empowers businesses with fast and reliable performance for all their key business workloads.

This collaboration with software providers has also enabled persistent memory to generate another key business benefit—cost savings. Compared to servers using solid-state drives, HPE Scalable Persistent Memory reduces CPU core requirements by as much as 50%, which significantly lowers licensing costs for software that are based on the number of CPU cores.

To find out more about how persistent memory can play a key role in data center modernization by helping your business gain greater IT agility that enhances your competitive edge, identifies areas to cut costs, and uncovers new ways to drive revenue, read more about scalable Persistent Memory. We also invite you to view a white paper that analyzes the cost and performance benefits of deploying Microsoft SQL Server on HPE Persistent Memory.

Related link:  IT on the fly

0 Kudos
About the Author


Michael Pratt’s passion is helping customers go further. His job is making products that make servers go further. He spends his days connecting the two.