#Gate广场中秋团圆局 #AMD预计2030年AI市场达2万亿美元 The AI market (especially AI computing infrastructure) will continue to rise in the foreseeable future. AMD expects the overall computing market (including AI accelerators, CPUs, and more) to reach $2 trillion by 2030. This forecast reflects the industry's long-term positive trend, but the slope of the market's growth and its ultimate scale will be influenced by multiple factors, including supply and demand, technological iteration, and the macro environment.
I. Core drivers of market growth
1. Structural shift in AI workloads (explosion of inference and agents): AI computing demand is shifting from “training” to “inference,” with inference tasks expected to account for more than 60%. Meanwhile, the adoption of agentic AI is placing higher demands on CPU scheduling, memory, and system coordination, driving the simultaneous expansion of full-stack computing (CPU+GPU+network) and raising the ceiling for overall computing demand.
2. Large-scale deployment of AI applications: The rapid adoption of large models, enterprise AI applications, AI agents, and other use cases continues to drive up computing demand across industries. Capital expenditure (Capex) on AI infrastructure (AI data centers) remains high, providing sustained growth momentum for the computing market.
3. Pressure from tight computing supply: Current AI computing supply (such as advanced packaging, HBM memory, optical interconnects, and power) still faces bottlenecks. The scarcity of computing resources has kept hardware prices relatively firm and supported pricing for computing services (such as GPU hours), further underpinning the expansion of the computing market.
II. Constraints and uncertainties on market growth
1. Balance of returns on computing investment (ROI): AI computing infrastructure requires massive investment. If the pace of commercial monetization of AI applications is slower than expected, or declining computing costs reduce the value of each unit of computing, some providers' willingness to continue investing may be affected.
2. Technological iteration and the competitive landscape: Rapid iteration of AI chips (such as GPUs and ASICs) and intense competition among vendors could cause some older computing resources to depreciate, affecting linear market growth. Meanwhile, the rise of in-house chips developed by some hyperscalers (such as Google, Meta, and AWS) could affect the allocation of incremental growth in the general-purpose computing market.
3. Macro and policy environment: Tightening global AI regulation, geopolitical restrictions on advanced chip exports, and macroeconomic disruptions to technology capital expenditure will all bring uncertainty to the long-term growth of the AI market.

In short, the growth cycle of the AI market (computing infrastructure) is far from over, but the market will shift from “broad-based growth” to “structural divergence.” Segments with technological barriers, the ability to provide highly cost-effective computing (low TCO), and the capacity to achieve commercial monetization quickly will gain greater room for growth.$AMD