Artificial Intelligence (AI)-optimized Infrastructure as a Service (IaaS) is fast becoming the next big growth driver in AI infrastructure, according to a new Gartner report. The research firm projects end-user spending on AI-optimized IaaS to surge 146% by the end of 2025, reaching $18.3 billion, and nearly doubling again to $37.5 billion in 2026.
The AI-optimized IaaS segment encompasses high-performance computing (HPC) resources such as GPUs, application-specific integrated circuits (ASICs), and other AI accelerators built to handle large-scale AI processing workloads.
“Traditional IaaS is maturing, however, AI-optimized IaaS spending growth projections are higher than that of traditional IaaS over the next five years,” said Hardeep Singh, Principal Analyst at Gartner. “As organizations expand their use of AI and GenAI, they will require specialized infrastructure — GPUs, TPUs, or other AI ASICs, high-speed networking, and optimized storage — for fast parallel processing and data movement. CPU-based IaaS will face significant challenges in meeting these demands.”
Gartner’s analysis highlights that inference workloads—which power real-time AI applications—are rapidly outpacing training workloads in driving demand. Spending on inference-focused applications is expected to rise from $9.2 billion in 2025 to $20.6 billion in 2026, accounting for 55% of total AI-optimized IaaS spending. By 2029, inference workloads are expected to represent more than 65% of the category.
“Unlike training, which requires large-scale compute cycles during model development, inference happens continuously—fueling real-time applications such as chatbots, recommendation engines, fraud detection systems, and industry-specific AI services,” Singh added.
According to Gartner, as AI becomes embedded across business functions, AI-optimized cloud infrastructure will underpin the next wave of enterprise-scale AI adoption, transforming how organizations design, deploy, and manage intelligent systems.
