Market scale is projected to expand to $66 billion (approximately $93 trillion KRW) by 2027.

Hardeep Singh, Senior Analyst at Gartner, stated, "With large language model (LLM) training infrastructure demand continuing, rapid AI adoption across enterprise applications and business processes overall is driving market growth."

With the proliferation of agentic AI, computing intensity in multi-stage autonomous processing is increasing, and resource consumption is being reorganized around inference.

Global inference spending this year is expected to reach $23.3 billion (approximately $33 trillion KRW), surpassing training-related spending of $19 billion (approximately $27 trillion KRW) for the first time.

The share of inference in total AI-optimized IaaS spending is projected to expand from 55% in 2026 to 59% in 2027.

Analyst Singh explained, "As enterprises transition from model development stages to large-scale commercial deployment, fine-tuned models and domain-specific models (DSM) are being rapidly integrated into customer response and operational systems. These systems require continuous real-time execution rather than one-time training, accelerating cloud consumption and continuously generating demand for AI-optimized infrastructure."

The gap with the broader IaaS market is also evident.

In 2026, while the overall IaaS market growth rate is 29.3%, AI-optimized IaaS is 96.4%, more than three times higher.

△AI-optimized IaaS 2025 spending $21.529 billion (growth rate 180.0%) △2026 $42.276 billion (growth rate 96.4%) △2027 $66.143 billion (growth rate 56.5%).