AI infrastructure spending to hit USD $37.5bn by 2026, says Gartner
Gartner forecasts that global end-user spending on AI-optimised infrastructure as a service (IaaS) will reach USD $18.3 billion by the end of 2025 and is expected to further increase to USD $37.5 billion in 2026.
The growth of AI-optimised IaaS, which encompasses expenditure on high-performance computing (HPC) resources such as graphics processing units (GPUs), application-specific integrated circuits (ASICs), and other accelerators for AI, is increasingly seen as a new engine for expansion in the broader AI infrastructure market.
Growth outpaces traditional IaaS
According to Gartner, the expansion rate of AI-optimised IaaS is projected to sharply outpace that of traditional IaaS solutions over the next five years, marking a significant shift in enterprise infrastructure priorities. The firm predicts that end-user spending in this segment will rise by 146% by the end of 2025.
"Traditional IaaS is maturing, however, AI-optimised IaaS spending growth projections are higher than that of traditional IaaS over the next five years," said Hardeep Singh, Principal Analyst at Gartner. "As organisations expand their use of AI and GenAI, they will need specialised infrastructure such as GPUs, tensor processing units (TPUs) or other AI ASICs, high-speed networking and optimised storage for fast parallel processing and data movement. As such, traditional central processing unit (CPU)-based IaaS will face significant challenges in meeting these demands."
The move towards AI-optimised solutions comes as enterprise requirements for computational resources grow alongside widespread adoption of advanced AI capabilities and generative AI models. These demands mean traditional CPU-based IaaS offerings may increasingly struggle to address the needs of businesses scaling up their AI initiatives.
Inference workloads driving demand
Gartner's analysis highlights that inferencing workloads, rather than training-intensive applications, will become the principal driver of spending in the AI-optimised IaaS market. Projections indicate that by 2026, spending on inference-focused applications will rise to USD $20.6 billion from USD $9.2 billion in 2025. In the same period, 55% of total AI-optimised IaaS spending is expected to be allocated to inference workloads, a figure forecast to increase to over 65% by 2029.
"Unlike training which involves intensive, large-scale compute cycles that occur during model development and ongoing updates, inference happens continuously - powering real-time applications such as chatbots, recommendation engines, fraud detection systems and industry-specific applications," said Singh.
This shift reflects the evolution of AI deployment patterns within organisations, with a growing emphasis on real-time inference to support a variety of applications across sectors. These include consumer-facing tools like chatbots and recommendation engines, as well as industry-specific systems such as fraud detection platforms.
Gartner's findings also indicate that as AI adoption becomes further embedded across a range of business processes, the infrastructure required needs to be tailored to support high-speed, parallel operations and rapid data movement, which favour accelerator-based compute platforms over traditional architectures.
Wider industry implications
The anticipated growth in AI-optimised IaaS spending is likely to have significant implications for technology providers and end users alike, prompting adjustments to cloud strategies and infrastructure investments. As inference workloads become more pervasive and demands for specialised hardware increase, service providers and enterprises will need to evaluate requirements for agility, scalability, and performance in supporting AI-driven applications.
