Generative AI will continue to shape cloud in 2024 — and users stand to benefit.
Cloud providers will likely bow to demands for cost standardization and data security improvements, according to a Monday research report from analyst firm Forrester.
Stricter cloud sovereignty and data center sustainability regulations coupled with potential chip shortages will drive hyperscalers to adapt or risk losing business, the report said.
“The rise of AI-focused cloud providers with ready-to-go GPUs will grab some workloads from big cloud providers,” Forrester Principal Analyst Lee Sustar told CIO Dive, pointing to Oracle’s AI-optimized Supercluster infrastructure and its recently expanded partnership with GPU manufacturer Nvidia.
The generative AI wave raised the specter of disruption in cloud, as the three largest hyperscalers have raced to deploy the technology across their platforms in search engines, coding assistants and on multimodel marketplaces.
“AI is restating the value proposition of cloud,” Sustar said. “There is the AI in the cloud — where you get GPUs and train models — and the AI of the cloud, in which platform operations are increasingly automated with the addition of AI services as well as AI-enhanced versions of services already in use.”
Intent on capturing both value streams as enterprise use grows, the hyperscalers will announce 30 new cloud regions globally in 2024 and introduce as-a-service prompt engineering support to ease generative AI applications, according to Forrester projections.
Additional data centers will help assuage data sovereignty concerns and mitigate service outage disruptions. Forrester anticipates that each of the three major cloud providers will have “announced or launched at least two geographically separated regions in every significant market to stay ahead of sovereignty requirements, climate change challenges and risks connected to geopolitical tensions.”
Forrester is less sanguine about the prospects for prompt engineering services, as most companies will prefer to leverage internal subject-matter experts and data science capabilities to cultivate talent.
To rationalize billing, hyperscalers will adopt standards developed by the FinOps Foundation through its FinOps Open Cost and Usage Specification. A new version of the FOCUS framework, which is expected before the end of the year, “will compel cloud providers to align on cost reporting to give customers a vendor-neutral multicloud view of their resources,” the report said.
Two of the big-three cloud companies, Google Cloud and Microsoft, have already signed up for the project. Forrester expects AWS to join the effort in 2024.
Standardizing cloud pricing isn’t likely to decrease overall cloud spend, which is expected to rise in 2024, driven by AI adoption, price increases and ongoing enterprise modernization.
“The constituency for AI in most organizations extends far beyond cloud infrastructure, so if the cloud providers are the main providers of AI capabilities, organizations may boost their overall cloud spend,” Sustar said. “At the same time, they will continue their efforts over the last couple of years to eliminate redundancies and consolidate cloud IT where possible.”