This voice is automatically generated. Please let us know if you have any feedback.
Diving overview:
- The surge in demand for automation is expected to triple most companies’ AI infrastructure budgets and increase investment in and ownership of the physical infrastructure needed to power AI by 2028. Deloitte report It was published last week. The company surveyed 515 large U.S. companies in December for the report.
- Almost half of respondents said they have 30 or more AI pilots in development. Deloitte predicts that by 2028, nearly 70% of enterprises will be running this many AI proofs of concept.
- Respondents said they use a mix of closed, open, derived, and packaged SaaS models in their AI toolkits, with closed models leading slightly. According to the report, there is no clear direction as to which model or combination of models will win in the coming years.
Dive Insight:
As the demand for AI accelerates, companies are working to build a blended portfolio of AI architectures. This recipe includes hyperscaler models and services, as well as a combination of public cloud and your own on-premises infrastructure.
Demand for AI infrastructure is universal – the neocloud market is set to reach $400 billion in revenue by 2031as big tech companies like Meta make deals to acquire more computing. this week, human signed a new contract Partnership between Google and Broadcom To add a few gigawatts TPU capacity from 2027 onwards As customer demand accelerates. The AI model maker also signed a multi-year agreement with CoreWeave for computing access.
The only way some companies can scale AI at their desired pace in a crowded resource market is by providing their own infrastructure. Chris Thomas, U.S. cloud strategic growth services leader at Deloitte, told CIO Dive.
“This is nearly doubling overall in three years, driven primarily by token volumes doubling and tripling,” he said. “Currently, we are creating workloads that cannot be cost-effectively supported by the public cloud alone.”
Investments in these projects are significant, with some companies predicting they will spend nearly four times as much on AI infrastructure by 2028. This is a shift from traditional IT spending that was planned as a one-time modernization effort. The report says IT departments are moving to more sustainable spending year-over-year.
CIOs may need to work more closely with other financial decision makers within the enterprise to demonstrate the value of infrastructure spending. thomas Said. Companies will need to make strategic choices to prioritize spending on AI. Some may choose to shift operational costs to capital expenditures by owning the infrastructure instead of renting cloud capacity.
“No matter what industry you’re in, the line between business and technology is blurring,” Thomas said. “Technology decisions must be made in collaboration with the business, and given the size of the budget required, there are few options.”
