- The optimistic performance guidance released by Advanced Micro Devices (AMD:US) significantly boosted market sentiment, driving its pre-market stock price up nearly 18%, and leading to a broad rally among core chip designers and manufacturers such as Intel (INTC:US), Arm Holdings (ARM:US), and Qualcomm (QCOM:US).
- With the widespread application of Agentic AI systems, the demand for computing power is extending from large model training to practical application inference. AMD expects the potential target market for server CPUs to have a compound annual growth rate of over 35% by 2030.
- Super Micro Computer (SMCI:US) provided better-than-expected guidance for fourth-quarter revenue and profit, with its stock price surging nearly 19%. The strong demand for customized AI servers temporarily alleviated market concerns about governance risks due to the U.S. Department of Justice's export control investigation.
Marginal Extension of Computing Power Investment Logic to the Inference End
The U.S. semiconductor sector exhibited significant sector rotation characteristics catalyzed by AMD's performance guidance. The market, originally highly concentrated on the investment logic of graphics processing units (GPUs), is now spilling over into the fields of central processing units (CPUs) and customized chips. Agentic AI requires systems to have autonomous execution capabilities and more complex logical judgments, making computing power architecture more reliant on the general computing and scheduling capabilities of CPUs during the actual deployment phase. AMD's substantial upward revision of the long-term growth forecast for the server CPU target market from 18% to over 35% directly reflects this technological evolution trend. This shift towards the inference field provides substantial performance growth options for non-GPU core targets, prompting funds to seek valuation gaps and catch-up targets across the entire semiconductor industry chain.
Repricing of Valuation Premiums and Market Share
In terms of secondary market pricing, AMD's cumulative gain of over 66% this year significantly outperformed Nvidia's (NVDA:US) approximately 5% gain over the same period. However, this strong relative performance has also elevated its valuation multiples. Currently, AMD's price-to-earnings ratio is about 39.66 times, not only far exceeding its historical average over the past five years but also nearly double Nvidia's approximately 21 times price-to-earnings ratio. This high premium state indicates that the market has already priced in expectations for AMD to capture a higher market share in the server CPU and emerging AI inference markets. Hargreaves Lansdown's analysis points out that as AI workload requirements become more complex, AMD's investment logic is no longer limited to challenging Nvidia's GPU dominance but is focused on a broader heterogeneous computing landscape. If subsequent performance fails to meet expectations, the high valuation multiples will face correction pressure.
Diversified Evolution of Server Architecture
In addition to core chipsets, memory and network infrastructure also benefit from this round of AI spending diffusion. Micron Technology (MU:US) rose 6.4%, while Marvell Technology (MRVL:US) also recorded a 1.7% increase. The simultaneous rise of these assets confirms that data centers are undergoing a deep architectural restructuring. High-bandwidth memory (HBM) and high-speed interconnect chips are key bottlenecks supporting the operation of multimodal large models and Agentic AI. Institutional investors are beginning to view the entire AI infrastructure as an indivisible hardware ecosystem, where the performance bottleneck of a single component will directly limit the throughput of the entire system. Therefore, chip companies with comprehensive platform integration capabilities or monopoly advantages in key niche tracks continue to receive premium allocations from long-term funds.
Compliance Risks and Fundamental Struggles in the Supply Chain
In the system integration segment, Super Micro Computer's (SMCI:US) market performance revealed the intense struggle between strong fundamentals and compliance risks. Despite the cloud of a U.S. Department of Justice investigation into illegal chip exports to China, the company recorded a nearly 19% gain in a single day due to its far better-than-expected financial guidance for the fourth quarter. This shows that in the current state of extreme hunger for high-performance customized AI servers among data center operators and startups, the market is more inclined to assign higher pricing weight to order visibility and capacity expansion. CEO Charles Liang's statement about significant expansion of production bases in Taiwan, Malaysia, and the Netherlands further reinforced performance certainty. However, JPMorgan's warning also indicates that corporate governance and geopolitical regulatory compliance remain long-term risks hanging over the company, and once regulatory measures are implemented, they could trigger severe supply chain disruptions.
