AI adoption to double storage calls for in simply 3 years, Seagate survey finds

In a nutshell: A brand new world survey commissioned by storage large Seagate reveals that companies anticipate their knowledge storage wants will double inside the subsequent three years, pushed largely by the adoption of AI and machine studying. The survey, which included over 1,000 companies throughout varied industries, discovered {that a} whopping 72% have already jumped on the AI practice, with one other 28% planning to deploy it inside the subsequent three years.

The info deluge is principally being dealt with by the cloud. The survey signifies that 65% of all AI-related knowledge is already being saved within the cloud as of 2024, and that proportion is projected to climb to 69% by 2028 as extra workloads shift off-premises.

Amongst firms utilizing over 100 petabytes of storage, 87% are storing their AI mannequin coaching checkpoints within the cloud or a hybrid cloud/on-premises setup. These checkpoints, which permit fashions to choose up the place they left off, are being generated at a blistering tempo: 28% of firms stated they create new checkpoints each single day, whereas 43% do it on a weekly foundation. It is no marvel almost half of the businesses surveyed (46%) consider their present storage strategies will not be sufficient to quench AI’s thirst for knowledge within the coming years.

Maintaining all these checkpoints round can be turning into a precedence. A full 90% of respondents stated they consider retaining numerous historic coaching knowledge results in higher AI outcomes over time. 32% of firms creating each day checkpoints are holding on to that knowledge for over a 12 months, whereas 29% maintain it for 6-12 months.

To attempt to get forward of any knowledge bottlenecks, companies are embracing quite a lot of methods: 61% are increasing their utilization of scalable cloud storage options; 56% have carried out enhanced knowledge administration software program; 55% have upgraded their on-prem infrastructure; and 49% are turning to knowledge compression to squeeze extra into their present storage.

With AI proving to be an exponential knowledge multiplier, companies don’t have any selection however to make scaling storage capability a high precedence over the approaching years. After all, cloud will possible proceed gobbling up extra of the storage load, but it surely appears on-premises optimization and hybrid setups will stay important too.