Export control pressure on Nvidia and China has moved Washington to the center of the AI chip supply chain, and a screen published by Yahoo! Finance on November 15, 2026 identifies three US-listed stocks positioned to benefit from onshore AI compute buildouts: Oracle, SharonAI Holdings, and Amphenol.
The policy shift is doing more than reshuffling chip logistics. It is forcing a rearchitecture of where AI computing physically lives. Compute, power, racks, and real estate are migrating onshore, and the investment case follows the infrastructure layer.
Oracle: gigawatt-scale compute as a national asset
Oracle generates roughly US$62.8 billion annually from cloud and software, US$5.8 billion from services, and US$3.2 billion from hardware, with a market capitalization of US$390.8 billion. Its US cloud regions already function as compliant, high intensity compute hubs for corporate and government workloads.
According to TickerTickle, an investor in the Simply Wall St Community cited by Yahoo! Finance, Oracle's current buildout is changing its competitive position against larger hyperscale rivals.
Oracle is building gigawatt-scale data centers; one massive 1,100-acre campus in Abilene, Texas, featuring eight buildings, is dedicated initially to OpenAI workloads, and once finalized will be the largest supercomputer ever constructed.
Abilene is not a data center. It is a single supercluster, networked through massive interconnections. Treat it as protocol infrastructure: capacity that, once live, either converts into durable demand or sits as stranded capex. The Motley Fool recently framed the same tradeoff, noting Oracle remains a high-stakes bet that AI demand will last for years, with shares down about 34% this year offering a lower entry point for believers.
SharonAI Holdings: pure-play onshore compute
SharonAI Holdings runs high performance AI data center infrastructure and cloud GPU platforms for AI-focused customers. It currently generates about US$3 million in revenue from high performance compute services, entirely within the United States, against a US$808 million market capitalization.
The thesis is a single relationship: VAST Data. Per the analyst narrative on Simply Wall St, SharonAI is deploying 600 petabytes of the VAST AI Operating System across its infrastructure, building a recurring storage and data services base that scales with GPU deployments and carries a richer margin profile than pure compute alone.
At this revenue scale, the model is optionality. The variable to watch is fill rate: how quickly higher value services absorb available capacity.
Amphenol: the physical interconnect layer
Amphenol supplies the high speed electrical, power, and fiber optic connectors that link nearly every major architecture in high density US data centers. Revenue runs about US$20.1 billion from Communications Solutions, US$7.4 billion from Harsh Environment Solutions, and US$6.0 billion from Interconnect and Sensor Systems, with roughly US$12.5 billion generated in the US. Market cap: US$193.2 billion.
Investor isidrohg, writing in the Simply Wall St Community, points to the order book as the signal.
Book-to-bill holding above 1.0. At 1.23 with record orders, this is the cleanest forward indicator in my whole portfolio.
A book-to-bill of 1.23 means orders outpace shipments by a wide margin. It is a measurable, falsifiable indicator, which is more than most AI narratives offer.
Why onshore positioning matters now
Export controls have turned geography into a compliance variable. Workloads that must satisfy US regulatory boundaries need US-resident compute, and that demand flows to whoever owns the power, the racks, and the fiber. Benzinga reported that HSBC recently named Oracle among 11 AI stocks to buy on the dip, alongside data center suppliers GE Vernova, Vertiv, and Bloom Energy, confirming that institutional desks are pricing the same shift.
The Yahoo! Finance screen surfaced 54 additional companies with similar profiles beyond the three highlighted here. The infrastructure layer of the AI economy is broad, and the buildout is only in its first cycle.
FAQ
Why are US AI data center stocks attracting attention in November 2026?
Export control measures around Nvidia and China are pushing AI computing onshore, and investors are repositioning around the companies that supply compute, power, interconnects, and real estate inside the US.
What is Oracle building in Abilene, Texas?
A 1,100-acre campus with eight buildings, initially dedicated to OpenAI workloads, designed to function as a single networked supercluster and described as the largest supercomputer ever constructed once complete.
How large is Amphenol's book-to-bill ratio?
According to an investor narrative on Simply Wall St, Amphenol's book-to-bill stands at 1.23 with record orders, indicating orders significantly outpacing shipments.
