Nvidia Stalls. Intel, AMD Start Doubling Rallies as AI Agents Lead Dramatic Shift in Chip Landscape

TradingKey - In the recent semiconductor market rally, NVIDIA, which once reigned supreme, (NVDA) appears to have been sidelined, as capital flows into once long-neglected Intel (INTC) , AMD and Micron (MU) and other hardware companies.

Year-to-date, "market king" NVIDIA has risen only 13.96%, slightly outperforming the Nasdaq's 12.96% gain, while Intel, AMD, and Micron have seen their share prices double this year, with Intel surging over 100% in the past month.

This situation corroborates the potential transfer of power in the AI sector proposed by Mizuho (MFG) analyst Jordan Klein. This shift is fundamentally driven by the pivot of AI applications toward AI Agents, which has created entirely new hardware demands.

Capital Exits Nvidia as Intel and Micron Take Up the Mantle in Market Rally

Currently, investors are betting on the long-term sustainability of the AI bull market, anticipating that AI data centers will require a broader variety of advanced hardware components in the future; this trend is reflected in stock movements as multiple hardware companies surge simultaneously.

Micron's market capitalization surpassed the $800 billion mark for the first time this week, with its stock price surging more than 750% over the past year. CEO Sanjay Mehrotra stated in March that due to supply constraints, major customers are only receiving "50% to two-thirds of their demand." Klein noted that this occurs when a market rapidly enters a raw material shortage, causing selling prices to skyrocket while costs rise "only slightly"—investors overweight on the cyclical upturn in memory stand to gain substantial returns. Currently, the memory market is dominated by Micron and South Korean firms Samsung and SK Hynix, operating as a seller's market.

Beyond the demand for memory, demand for CPUs (central processing units) has also been climbing rapidly, primarily driven by the higher CPU ratios required in the AI Agent era. When model developers like OpenAI and Anthropic first emerged, the role of CPUs in AI data centers was largely overlooked; tech giants, particularly hyperscale cloud service providers, were heavily purchasing Nvidia GPUs, leaving the CPU market neglected. The reassessment of the CPU market's value is the core reason for the rapid rise in CPU-related stocks.

CPU giant AMD's earnings report released this week highlights this trend. In addition to earnings, revenue, and guidance significantly exceeding expectations, CEO Lisa Su predicted that the server CPU market will grow by 35% over the next three to five years, up from the 18% forecast the company made last November. Bank of America (BAC) estimates that the data center CPU market size could grow from $27 billion in 2025 to $60 billion by 2030, more than doubling.

AMD's rival Intel also recorded its best monthly performance ever in April, gaining over 100%, and continued its strong momentum in early May with a 33% increase. While the company led AMD in the CPU market for years, it repeatedly missed growth opportunities in AI and lagged behind. Since the U.S. government took a stake in Intel last year, the company's business operations have begun to recover.

This Tuesday, Bloomberg reported that Intel is in talks with Apple (AAPL) to manufacture main processors for the latter's U.S. devices, sending Intel's stock price soaring 13%; on Friday, a Wall Street Journal report indicated that Intel has reached an agreement with Apple to produce some processors for Apple devices, causing Intel shares to rise another 14%.

From Nvidia GPUs to Intel CPUs: Which will be the next to surge?

As AI Agents become the mainstream of AI applications, capital flows in U.S. equities have shifted, and the semiconductor sector has entered a new phase of structural divergence. Currently, investors are no longer blindly pursuing broad AI investments but are looking for targets that truly solve AI bottlenecks, such as the HBM oligopoly of Micron, Samsung, and SK Hynix, and CPU giants AMD and Intel. This explains Nvidia's lackluster performance so far this year, while former "supporting actors" have taken center stage.

As the AI boom becomes long-term, the market will continue to seek value in areas outside the semiconductor sector. For instance, beyond CPU and memory hardware, Corning, which specializes in fiber optic technology, (GLW) has surged more than 100% this year. The company signed a major partnership with Nvidia this week to build three new U.S. factories dedicated to Nvidia's optical technology requirements, a move seen as pivotal in Nvidia's transition from copper to fiber optic cabling to support its rack-level system expansion.

Beyond companies involved in chip materials, tech firms that service data centers may also attract interest. In this space, Vertiv (VRT) is the global leader in precision cooling and liquid cooling, and Eaton (ETN) is a power management giant providing stable grid upgrade solutions. These companies are all potential beneficiaries of an AI supercycle.

However, some analysts warn that current conditions mirror the late-1990s dot-com boom, which was followed by a protracted market crash. BTIG analyst Jonathan Krinksy noted in a recent report that the semiconductor industry's gains resemble the 1999 tech bubble. He cautioned that the Philadelphia Semiconductor Index, a key benchmark, could see a 25% to 30% correction, despite rising 66% year-to-date. Krinksy suggested that the current semiconductor trajectory might actually be more extreme than that of the dot-com era.