Key Takeaways:
AI infrastructure spending is rotating beyond semiconductor giants into Asian suppliers of servers, cooling systems, and power equipment as the next phase of capital deployment takes shape.
As SpaceX, OpenAI, and Anthropic prepare to raise tens of billions of dollars in fresh capital, investors are shifting focus from chipmakers to the Asian supply chain companies building the physical infrastructure that powers artificial intelligence, according to a Bloomberg report. The three companies could collectively raise tens of billions, adding to more than $750 billion already committed by major technology firms to AI-related capital expenditures.
"The first wave of AI spending went to the picks-and-shovels — the GPU makers. The second wave flows to everything that supports them," said a Bloomberg report citing market participants. Investors flagged concerns that valuations of leading semiconductor stocks have become stretched after strong gains driven by demand for AI chips and data-center equipment.
Attention is turning to companies producing electronic components, advanced packaging materials, cooling systems, power equipment, and server-related products. Samsung Electro-Mechanics and Japan's Ibiden have benefited from growing investor interest in suppliers tied to AI infrastructure. Taiwan's Hon Hai Precision Industry, Quanta Computer, and MediaTek were also cited as companies positioned to benefit from continued AI-related spending.
The shift comes as Nvidia, the dominant beneficiary of the first AI spending wave, reported its strongest quarter in company history. Data-center sales hit $75.2 billion in the most recent quarter, while total revenue climbed 85% year over year to $81.6 billion. Adjusted earnings per share rose 140% to $1.87. Yet Nvidia's trailing price-to-earnings ratio of 32 sits well below its five-year median of roughly 61, a gap that some analysts argue leaves room for further upside even as the investment thesis broadens.
The rapid expansion of data centers has intensified focus on electricity generation and transmission, prompting investors to look at companies involved in power equipment, nuclear energy, and renewable energy projects. In South Korea, HD Hyundai Energy Solutions and Daewoo Engineering & Construction have been among the stronger-performing stocks this year as investors position for rising energy demand linked to AI development.
Nvidia itself is deepening its physical and financial roots in Asia. The company plans to ramp its annual spending in Taiwan to $150 billion, reflecting its interdependence with fabs at Taiwan Semiconductor Manufacturing Co., packaging specialists at ASE, and chip designers such as MediaTek. A new corporate headquarters in Taipei is slated to break ground by mid-2027 and become operational by 2030.
Some investors are also targeting companies involved in robotics and autonomous systems, areas often referred to as "physical AI," which have gained momentum through partnerships with companies such as Nvidia. The broadening of the AI investment thesis suggests the infrastructure buildout will support demand across a wider range of companies than the semiconductor leaders that dominated the initial phase of the rally.
For investors, the rotation raises a question of valuation. Nvidia trades at 32 times trailing earnings, a discount to its five-year median, while many of the Asian supply chain beneficiaries trade at lower multiples with less AI premium priced in. If the $750 billion in committed CapEx continues to flow at current rates, the companies building the physical layer of AI infrastructure may capture a growing share of that spending in the years ahead.
This article is for informational purposes only and does not constitute investment advice.
