- Data & Analytics
How AI tools are changing the need to build more elasticity in enterprise data storage
Artificial intelligence, machine learning and deep learning are revolutionizing how industry and higher education institutions are analyzing big data. But they are also forcing CIOs to rethink how they store and manage vast pools of data, says a new report produced by MIT Technology Review and underwritten by Pure Storage.
The report, “Modern Storage Accelerates Data Insights, Speeding Innovation,” makes the case for organizations to modernize their IT infrastructure to be more elastic, and it highlights key considerations to create elasticity effectively. This will give them flexibility to manage quickly evolving applications and data types.
Preeminent AI expert Andrew Ng believes advanced deep learning will be transformative across all industries in the near future. Ng cites, as an example, the rise of voice-interface devices such as Amazon Alexa, which require high-quality speech recognition.
Data is critically important to next generation research and innovations. Businesses are already considering storage as an essential piece of the equation when thinking about AI, and higher education institutions must do the same, the report says.
A Pure Storage executive in the report noted that as the nature of data changes, organizations need a platform built to deliver performance for modern faster processing.
Organizations still using legacy disk-based storage systems are likely to realize those systems won’t be able to deliver data fast enough to harness the potential of deep learning applications, the report says.
With disk-based storage, IT departments tend to overprovision their systems to hedge against the inherent weaknesses of mechanical storage, says Scott Sinclair, senior analyst at Enterprise Strategy Group.
“[Overprovisioning] causes problems — running out of space, consuming lots of power and then needing more power to cool all the heat. And there are many more components that can fail,” he explains.
However, flash storage alleviates the problems associated with spinning disk storage because its parallel-access architecture meets the requirements of simultaneous scale and performance.
All-flash storage infrastructure, like Pure Storage’s FlashBlade for instance, is designed from the ground up to meet the demands of advanced data analytics.
When considering the prolific use of modern devices such as smartphones, iPods, iPads, and how these are all based on flash technology, enterprise storage is only just beginning to catch up, the report states.
Additionally, organizations are just at the beginning to apply deep learning technology more broadly and its possibilities are endless.
Read the full report for more information on preparing an elastic IT infrastructure to see the most benefits from the data analytics.
This article was produced by EdScoop for, and sponsored by, Pure Storage.