AI adoption is expected to drive exponential growth in data storage demand through 2028 as three in every five (61%) of infrastructure buyers who predominantly use cloud storage for AI data management expect storage requirements to at least double by 2028.
This comes from longer retention times of six months to forever, 73% using daily or weekly LLM checkpointing, and 80% deem data replication for AI very or moderately important.
These are findings from a study commissioned by Seagate Technology and conducted by Recon Analytics in November 2024. It involved a survey of 1,062 storage infrastructure buyers and decision makers from companies reporting greater than US$10 million in annual revenues and in excess of 50 TB of current storage capacity across 10 markets.
Each respondent included in the survey had to have already adopted AI or have plans to adopt AI in the next 3 years. Of those 1,062 respondents 72% are currently using AI and 28% plan to use AI in the next three years.
Findings also show that 95% of storage buyers, using AI or planning to, say they are taking measures to accommodate the growing storage requirements. They include 61% who are adopting more scalable storage, 56% who are implementing data management software, 49% who are using compression techniques and 55% who are upgrading existing storage infrastructure.
Recon Analytics’ research finds that wherever AI is adopted, existing storage practices will need to be upgraded to realise the full potential of AI.
The survey shows that cloud storage is expected to remain as the main storage vehicle for AI with 65% of data stored in the cloud versus in-house in 2024 and increasing to 69% by 2028.
Among respondents, 61% who predominately use cloud storage say their storage requirements will increase by over 100% over the next three years.
Also, 46% of respondents believe that existing data storage methods will not be enough to keep up with demand.
Additional data storage solutions are being adopted to manage the increasing file sizes and quantity generated by AI, including 61% expanding usage of cloud storage solutions, 55% upgrading existing infrastructure, 56% adopting enhanced data management software and 49% implementing data compression techniques.
Further, 25% of respondents said security was the most important component followed by 18% saying storage. Two-thirds (66%) of respondents ranked storage amongst their four most important infrastructure concerns, while 68% ranked security in the top four.
Nine in every 10 respondents who have adopted AI believe longer data retention improves the quality of AI outcomes. Among them, 93% claim data retention requirements have changed due to the implementation of AI and the ability to refine models including checkpoints.
The more data storage a company uses the more they see that longer retention times improve the quality of AI outcomes. The importance of data replication to a company’s AI data management strategy also increases the amount of storage a company uses.
More than half, (52%) of respondents who are currently using AI and who are also using more than 100 PB of storage, deem data replication improves AI outcomes as very important.
In addition, 73% of respondents say AI training is driving increased data storage as they are backing up their previously saved checkpointing data on a daily to weekly basis.
Of those respondents saving checkpoints daily (28% of respondents), 32% are retaining data for more than 12 months while 29% are retaining for six to 12 months.
Companies already using 100+PB of storage are saving and backing up checkpoints on a daily to weekly basis with 87% of them storing these checkpoints in the cloud or in a mix of HDD and SSD.
As AI use cases and adoption becomes more pervasive, Recon Analytics forecasts companies will see exponential growth in their storage requirements.
This will become even more evident when businesses move from their early AI trialing phase to being active AI users. Training LLMs, data replication and longer data retention periods, all key elements of an AI strategy, will require increased storage investments to be successful.