The AI infrastructure market is undergoing a significant evolution, transitioning from basic training capabilities to a system that supports a wider array of applications, including inference and enterprise deployment. This shift is driving a surge in capital expenditure as companies invest heavily to meet the growing demand for intelligent agents.
Leading this change, major OEMs such as Dell and Lenovo are seeing impressive growth in AI-related revenue. Dell recently reported a remarkable 757% year-over-year increase in AI server revenue, reaching $16.1 billion for the first fiscal quarter of FY2027. This strong performance has led the company to raise its full-year AI server revenue forecast from $50 billion to $60 billion. Similarly, Lenovo Group’s AI-related revenue jumped by 84% year-over-year, now making up 38% of its total revenue.
Evolving Demand for AI Infrastructure
The current wave of investment in AI infrastructure is not just a reaction to immediate needs; it reflects a long-term capital expenditure cycle. As C. C. Wei, CEO of Taiwan Semiconductor, pointed out, demand is shifting toward a model that combines training, inference, and Neocloud capabilities. This evolution is highlighted by Morgan Stanley's prediction that global data center construction costs will reach approximately $2.9 trillion by 2028, with capital expenditures in 2026 expected to hit $740 billion, a 69% increase year-over-year.
However, these advancements come with challenges. Industry experts have noted that rising compute costs, difficulties in quantifying business value, and the need to adapt to specific vertical scenarios could slow the widespread deployment of AI agents. The complexity of integrating various subsystems—such as high-performance CPUs and GPUs, advanced interconnects, and efficient cooling solutions—will determine which manufacturers can secure higher premiums in this competitive market.
The Rise of Intelligent Agents
The transition from generative AI to action-oriented agents marks a significant shift in how businesses leverage technology. Tools like Claude Code and the open-source framework Open Claw are enabling AI to serve as more than just chatbots; they are evolving into digital employees capable of managing workflows, executing tasks, and enhancing productivity. This shift is supported by insights from Alibaba's earnings call, where CEO Eddie Wu reported a notable increase in API call demand linked to advancements in AI coding capabilities.
NVIDIA is also making progress in this area with its Vera Rubin platform, designed to accelerate the development of AI agents. Jensen Huang, NVIDIA's CEO, designated 2026 as the year of Agent AI, signaling a pivotal moment for the sector. As these intelligent agents gain traction, the demand for Model-as-a-Service (MaaS) is projected to increase, further driving investment and innovation.
Challenges and Opportunities Ahead
Despite the excitement surrounding AI agents, several hurdles remain. Lu Yanxia from IDC China warns that most AI applications have not yet reached commercial viability. High computing costs and the difficulty of effectively measuring the value generated by AI tools are significant barriers to widespread adoption. Gartner predicts that over 40% of Agentic AI projects may be canceled by the end of 2027 due to these challenges.
Additionally, the organizational changes needed for successful AI integration involve not only technology but also a transformation of business processes and the development of talent that merges AI expertise with industry knowledge. Ji Xinhua, CEO of UCloud, stresses the importance of ensuring that employees effectively adopt AI tools, as realizing the value of such investments can be challenging.
The momentum behind AI infrastructure and intelligent agents is clear. While the environment is filled with challenges, the potential for growth and innovation is substantial. Companies like Dell and Lenovo demonstrate that the demand for advanced AI solutions is set to rise, provided the industry can navigate the complexities of deployment and integration. The race for effective AI applications is underway, requiring patience and strategic investment to seize the opportunities ahead.
Quick answers
What is driving the current demand for AI infrastructure?
The demand is shifting from training compute capacity to a broader focus on inference, Neocloud, and enterprise deployment.
Which companies are leading the growth in AI server revenue?
Dell and Lenovo are reporting significant increases in AI server revenue, with Dell's revenue increasing by 757% year-over-year.
What challenges does the AI industry face in deploying intelligent agents?
Key challenges include high computing costs, difficulties in measuring business value, and the need for adaptation to specific vertical scenarios.
How is NVIDIA contributing to the AI agent landscape?
NVIDIA has launched the Vera Rubin platform to support the development of AI agents, marking a significant step in the evolution of AI applications.
The stories that move AI & crypto markets — before the market reacts.
Free. 7am ET. Five stories. 62,400 readers.
