PA Consulting's Alastair McAulay, an IT expert, comments on AI and storage in an article in Computer Weekly.
The article goes on to look at how quickly AI systems need to process data. AI developers are building Graphics Processing Unit clusters which are the most effective ways to process the data and run algorithms at speed.
Alastair goes on to explain that academic and industrial high-performance computing (HPC) systems are typically run at very high utilisation rates because of their scarcity and cost.
He goes on to say that research institutes employ specialists to squeeze the last drop of performance from the hardware. In the enterprise, integration with existing data systems can be more important.
“We see judicious application of solid-state storage bringing a massive benefit,” says Alastair. “But it is more about what file system to use, how that is optimised, and whether any accelerators are needed to get the most from [off-the-shelf] storage hardware. They are putting most effort into file systems and managing data.”