
Anthropic’s blockbuster Nasdaq debut, with up to $81.3b raised on a roughly $1.5t valuation and nearly $900b of future compute commitments, has thrown a spotlight on the plumbing behind artificial intelligence. Capital is rushing to wherever the picks and shovels of this boom might live, and hesitation risks missing the next leg of AI infrastructure spending. This piece walks through three stocks that appear closely tied to that story.
The three stocks below are just a starting sample, since the full screen on AI infrastructure and enablers surfaced 72 more large, liquid companies with equally compelling narratives that are not covered here.
If you want to go straight to the source and identify your own high conviction angles across semiconductors, cloud platforms and data center providers, analyze the full AI Infrastructure and Enablers (Semiconductors, Cloud, Data Center Providers) screener.
Taiwan Semiconductor Manufacturing sits at the center of the AI infrastructure story in this screen, since it fabricates the advanced chips that power data centers, GPUs and cloud platforms. steingar, an investor in the Simply Wall St Community, argues that the real story is how AI reshapes the firm’s economics over time.
"AI and HPC wafer intensity: Advanced logic demand scales super linearly with compute complexity, not unit volumes."
What happens to Taiwan Semiconductor Manufacturing’s margins if that super charged compute demand meets one unresolved constraint in its vast buildout plans?
Taiwan Semiconductor Manufacturing is a pure foundry giant, fabricating CMOS logic and other advanced chips used in high performance computing, smartphones, cars and consumer electronics for clients around the world. Almost all of its NT$4,944,817 million revenue comes from the foundry segment, and its market cap sits around NT$56,402,893,875,000, putting it among the world’s largest listed companies.
If that constraint matters to your thesis, read the full narrative for Taiwan Semiconductor Manufacturing to see how AI wafer intensity, capital needs and pricing power might be decoupling.
Fabrinet plugs directly into the AI Infrastructure and Enablers theme by building the optical hardware that lets data centers move huge volumes of information between AI chips and servers. The analyst consensus Narrative on Simply Wall St frames that connection as a key earnings driver.
Fabrinet manufactures high speed optical networking equipment, with about US$5.1b of revenue tied to optical networking products, and has a market value of roughly US$11.2b.
"Surging AI related data center optics and DCI demand, with data center revenue at 51% of total and Q4 FY26 DCI products at more than a 1b annualized run rate, points to a multi year build out that could keep Fabrinet’s utilization and data center revenue elevated."
What happens to Fabrinet’s earnings power if a single pressure point changes how much of that AI data center build out actually runs through its factories?
If that pressure point matters to your thesis, read the full narrative for Fabrinet to see how AI optics demand, customer mix and pricing power could be shifting.
Suzhou TFC Optical Communication supplies optical sub-assemblies and packaging that link AI chips inside data centers, with almost all of its CN¥6.0b or so operating revenue coming from Optical Communication Device sales. The stock is valued at roughly CN¥119.4b, giving investors large-cap exposure to AI data traffic.
Suzhou TFC Optical Communication ties directly into AI infrastructure because its high speed optical engines, WDM components and packaging help move data between accelerators in AI computing centers and cloud facilities. Earnings, revenue and margins have all been growing alongside this demand. Investors still have to weigh what happens if a single unseen swing factor in AI capex timing hits that optical pipeline.
That hinge factor could be crucial, so read the 4 key rewards and 2 important warning signs (2 are major!) to see how Suzhou TFC Optical Communication’s AI exposure and risk profile really line up.
Fresh themes keep breaking out while yesterday’s ideas lose momentum. Before the best entry points attract too much attention and pricing power starts dropping, consider getting in early.
Have feedback on this article? Concerned about the content? Get in touch with us directly. Alternatively, email editorial-team@simplywallst.com