- MinIO's CTO thinks concerns about the privacy, security and cost of AI could drive a return to private cloud deployments
- Analysts disagreed on whether this will turn out to be true
- Gartner's Sid Nag called such concerns a "red herring"
Cloud growth is finally back on track after a wave of enterprise cost optimization and repatriation efforts resulted in something of a revenue slump. But could artificial intelligence (AI) trigger another private cloud push? MinIO thinks so.
Ugur Tigli, CTO of open-source data storage company MinIO, told Fierce that in the age of AI the move to a private cloud can offer wary enterprises more control over both their costs and their data.
It’s no secret that AI training requires boatloads of data. And Tigli said the AIOps pipeline requires data to be saved multiple times. When you factor in storage costs and egress fees, it adds up to a pretty penny pretty quickly.
Additionally, retrieval augmented generation (RAG) is emerging as a key technique for enterprises to get more accurate results from large language models by feeding their proprietary data into the model. But Tigli said large enterprises in regulated industries especially want to keep their data close to the chest. That very data, after all, is their secret sauce.
By moving to a co-location facility or on premises deployment, enterprises can own the full cloud stack and get a better grip on both cost and their data, Tigli argued. All it takes is some commodity hardware, three or four key pieces of software (usually a Kubernetes scheduler, network software and storage components), and voila.
MinIO, of course, has a horse in this race given it offers software-based object storage. But what do analysts think?
David Linthicum, former Chief Cloud Strategy Officer at Deloitte Consulting and founder of Linthicum Research, told Fierce that in many cases "on-premises resources are going to be more cost-efficient, and considering that AI systems cost 2 to 3 times the cost to build and run, enterprises are not automatically sold on public cloud providers for AI."
"On-premises systems are generally cheaper, can be more secure, and feel like they provide more control; enterprises are opting for them more than the cloud providers would like," he added. "I suspect that the cloud providers will see good growth around AI, but enterprises will be repatriating the platform options, considering the high cost of the cloud."
However, Sid Nag, Gartner VP Analyst, contended privacy and security concerns are a “red herring.” He argued that public cloud providers have “cracked the code on privacy and security” and pointed out that U.S. intelligence agencies run their operations on the public cloud.
Thus “there will not be a wave of repatriation,” Nag said. While there may be small language models that may not require the same scale of compute as the public cloud offers and could run on private, distributed clouds, the idea of a repatriation wave is “non-sequitur.”
It seems time will tell – and we’ll be watching!