GenAI fuels China’s $223B compute surge
China’s AI infrastructure is scaling fast as LLM workloads and enterprise adoption drive explosive growth in compute power.
Global AI server market to hit $158.7B in 2025, up from $125.1B in 2024, with generative AI servers accounting for nearly 30% of total spend.
China’s AI compute capacity to reach 2,781.9 EFLOPS by 2028, driven by rapid enterprise adoption, real-world deployments, and infrastructure diversification.
42% of Chinese enterprises are testing large models, with 17% already integrating them into production.
AI servers surge as generative workloads reshape the market
The global appetite for artificial intelligence infrastructure continues to accelerate.
According to IDC, the AI server market will reach $125.1 billion in 2024 and expand to $158.7 billion in 2025, eventually surpassing $222.7 billion by 2028.
Notably, generative AI workloads are gaining a growing share of the market, set to represent:
29.6% of AI server spend in 2025
37.7% by 2028
This shift signals a structural change in global compute investment, with cloud providers and enterprises alike rearchitecting infrastructure to meet massive model inference and training demands.