AI is having a reality check as Anthropic lines up a potential US$64b IPO and big tech commits US$26b of fresh capital to its non voting shares. Sentiment is flickering between excitement and nerves, which can throw prices around faster than the underlying businesses change. This article zeroes in on three stocks exposed to that Anthropic news and explains why some investors see opportunity while others stay cautious.
The three stocks below are just a sample. The full screen actually throws up 44 more large, liquid AI infrastructure and enabler companies with equally compelling stories that are not covered here.
If you want to move beyond headlines and start sizing up potential high conviction AI hardware and cloud plays for yourself, head straight to the AI Infrastructure & Enablers (Semiconductors and Cloud Hyperscalers) screener to identify, filter, and analyze them side by side.
Ultra Clean Holdings plugs directly into the AI Infrastructure & Enablers theme by building the subsystems and services that keep advanced chip factories running, and the analyst consensus Narrative on Simply Wall St focuses on how its expanding footprint might convert that demand into scale.
Ultra Clean Holdings develops outsourced subsystems, components, and cleaning services for semiconductor equipment makers and chip manufacturers, closely tied to AI driven fab investment. The business generated about US$2.2b from Products and US$282 million from Services, and the stock’s market value sits around US$2.2b.
Scaling of global manufacturing capacity, including 26,000 square feet of added cleanroom space in Malaysia and expansions in Singapore and the Czech Republic, targets support for a US$4b annualized revenue run rate by mid 2027 and positions Ultra Clean Holdings to translate higher industry volumes into operating leverage and margin expansion.
What happens to Ultra Clean Holdings margins if a single key assumption about future wafer fab equipment demand proves too optimistic or too modest?
That hinge point on wafer fab demand is exactly where the story gets interesting, and the full narrative for Ultra Clean Holdings shows how that sensitivity could accelerate or mask value.
AXT sits in the AI Infrastructure & Enablers screener because its compound semiconductor wafers feed into the high speed optics and communications hardware that keep AI data centers talking to each other. The analyst consensus Narrative on Simply Wall St frames that role around one key driver.
Accelerating AI data center optical buildouts are tying AXT to long-dated indium phosphide wafer supply agreements that include capacity reservations and prepayments of more than US$90 million through 2031, which can support multi year visibility in revenue from high value substrates.
The real swing factor is how one pressure on future capacity and funding shapes whether that visibility turns into the earnings profile investors are hoping for.
AXT is a Fremont based materials supplier that designs and manufactures compound semiconductor wafers, such as gallium arsenide and indium phosphide, used in high speed data links, 5G gear, sensors, and AI related optics. The business currently reports around US$171 million from Semiconductor Equipment and Services and carries a roughly US$2.8b market value.
That pressure point is exactly where full narrative for AXT helps you see whether accelerating AI optics demand could outweigh funding constraints or leave AXT’s earnings power stalling out.
STMicroelectronics designs chips that power data centers, edge devices, and AI hardware, which puts it squarely inside this AI Infrastructure & Enablers screen. The business generated about US$5.8b from Analog, MEMS & Sensors, US$4.3b from Embedded Processing, and carries a roughly €35.9b market value.
For AI focused readers, STMicroelectronics matters because it connects the high compute demand in data centers and at the edge with the less visible power, sensing, and control layers that Anthropic type workloads rely on. The analyst consensus Narrative on Simply Wall St spotlights how management is trying to reshape that manufacturing footprint for better long term earnings power.
The company's ongoing footprint reshaping and cost base resizing program has progressed to specific plans to close older 150 millimeter and 200 millimeter fabs around end 2027 or early 2028, with management attributing more than 400 basis points of potential gross margin uplift to this shift toward 300 millimeter and 200 millimeter production and therefore to higher future earnings power.
What that means for investors is that future AI heavy demand could interact with one crucial operational pivot that decides how much of it actually shows up in margins.
That operational pivot is where the story sharpens, and the full narrative for STMicroelectronics shows how AI heavy demand could turn footprint changes into accelerating, or even decoupling, earnings power.
Fresh breakout stories rarely stay quiet for long. Momentum can shift quickly, prices can move, and widely followed ideas can get crowded. Scan these under the radar lists while it matters and consider them before they become widely followed.
Have feedback on this article? Concerned about the content? Get in touch with us directly. Alternatively, email editorial-team@simplywallst.com