The global artificial intelligence (AI) infrastructure market is on a trajectory to reach $758 billion by 2029, with organizations more than doubling their investments in AI-related hardware, according to new data from market researcher IDC.
Spending on AI compute and storage hardware skyrocketed by 166% year-over-year in the second quarter of 2025, to $82 billion, reflecting sustained confidence in AI’s transformative potential across industries.
The spending splurge is primarily driven by investments in specialized server infrastructure. Cloud and shared environments dominate the landscape, accounting for 84% of total AI spending in the second quarter, with hyperscalers, cloud service providers, and digital service providers contributing 87% of quarterly expenditures.
Servers equipped with embedded accelerators — specialized chips designed to handle AI workloads — have become the infrastructure of choice, representing 92% of server-related AI spending and growing 207% year-over-year. IDC forecasts that accelerated servers will comprise more than 95% of AI server infrastructure spending by 2029, expanding at a 42% five-year compound annual growth rate.
“There is a distinct possibility that more AI-related investment will be announced in the coming years that will add to and extend the current mass deployment phase,” said Lidice Fernandez, group vice president of worldwide enterprise infrastructure trackers at IDC. She noted hyperscalers and cloud service providers will continue driving adoption, with AI-based research and education projects gaining prominence toward the end of the forecast period.
Geographically, the United States commands a dominant 76% share of global AI infrastructure spending, followed by China at 11.6%, Asia-Pacific at 6.9%, and Europe, Middle East, and Africa at 4.7%.
Over the next five years, China is expected to post the fastest growth rate at 41.5%, closely followed by the United States at 40.5%.
The forecast reflects a significant reassessment of GPU and accelerator demand, particularly in the U.S. IDC now expects the AI investment surge to continue through 2025 and into 2026, abandoning previous predictions of a slowdown based on expanding pipelines from major vendors and buyers.
Storage infrastructure is also experiencing robust growth, increasing 20.5% year-over-year in the second quarter as organizations grapple with massive datasets required for training AI models and storing checkpoints for inference operations. Cloud deployments account for 48% of AI storage spending, according to IDC.
