AI infrastructure stocks Lumentum, Celestica, Seagate beat Nvidia 2025

Nvidia has been the biggest infrastructure winner in the artificial intelligence boom, soaring in value by almost thirteenfold since the end of 2022 to a market cap of $4.6 trillion.

While Nvidia’s rally continued in 2025, investors betting on other AI data center plays made a lot more money over the past 12 months.

With four of the biggest technology companies projecting collective expenditures of $380 billion on data center and infrastructure build-outs this year, followed by an expected increase in the coming years, Wall Street has poured money into an assortment of vendors that are poised to reap the rewards.

Makers of memory, storage, fiber-optic cables, central processors and other types of enterprise hardware have rocketed in valued this year, driven by excitement around the AI craze.

Investors will be scrutinizing these five companies closely in 2026 now that lofty expectations have been built into their stock prices.

Lumentum, based in San Jose, California, makes switches, transceivers and other optical laser-based parts that are needed for fiber-optic cables. Customers have typically been telecommunications carriers and device makers like Apple, which previously used Lumentum parts in its FaceID sensor. 

But AI servers also need a lot of optical connections. Every graphics processing unit in a rack needs to be connected to every other GPU. Future AI systems will scale out, which requires optical connections from rack to rack. Eventually, entire data centers will need to be connected to each other with fiber-optic connections. 

Lumentum’s stock price has jumped 372% this year as of Wednesday’s close, lifting the company’s market cap past $28 billion. Sales surged 58% in the most recent quarter from a year earlier to $533 million. 

“Our growth is powered by AI demand spanning our laser chips and optical transceivers inside data centers, as well as the interconnected long-haul networks that link them,” Lumentum CEO Michael Hurlston said on an earnings call in November. He said 60% of the company’s sales now come from cloud and AI infrastructure.

Revenue is expected to rise 58% for the fiscal year ending in June, but analysts see a slowdown from there to growth of 32% and 15% in the next two years, according to LSEG.

Western Digital is one of three major hard drive manufacturers, along with Seagate and Toshiba. Shares of the 55-year-old company are up almost 300% this year.

In addition to computing power, AI companies need increasing amounts of space to store applications and other data. In short, data centers need hard drives.

While Western Digital makes solid-state hard drives, which use chips to store data, the company is best known for its hard disk drives, which use spinning discs to store terabytes or more of data.

“Data is the fuel that powers AI, and it is HDDs that provide the most reliable, scalable and cost-effective data storage solution,” CEO Irving Tan said in October on an earnings call. He cited an example of a hospital that’s using an AI that analyzes 7 billion images.

In the most recent quarter, revenue rose 27% to $2.82 billion. The company says that selling more storage for data centers will improve profitability because AI companies need larger, more expensive hard drives.

Revenue is expected to increase about 23% in fiscal 2026, with growth slowing to 13% in 2027.

In February, Western Digital spun out its flash business as Sandisk, which now has a market cap of about $35 billion, more than half the value of Western Digital.

Micron is one of three major memory producers, alongside Samsung and SK Hynix, but the only one based in the U.S.

Artificial intelligence servers need a lot of memory to store and process massive AI models. Chips from Nvidia or Advanced Micro Devices come with scores of gigabytes of the most advanced memory, called high-bandwidth memory. The chipmakers are taking all the memory production capacity, leading to a worldwide shortage and driving prices higher.

Micron blew away Wall Street estimates for sales and earnings in its quarterly report last week, and the stock is now up 241% for the year.

Sumit Sadana, Micron’s business chief, said the company was “more than sold out” of its memory chips. In December, it even shuttered its consumer-focused line of memory and solid-state drives to save supply for AI.

Analysts from Morgan Stanley said in a December note that Micron’s results showed the best revenue and profit upside in the “history of the U.S. semis industry” — aside from Nvidia.

Revenue is expected to almost double in the year ending in August, before dramatically slowing to 24% in fiscal 2027 and less than 1% in 2028, according to LSEG.

Seagate, founded nine years after Western Digital, is also benefiting from booming demand for storage. The stock is up 231% this year.

Sales rose 21% to $2.63 billion in the company’s fiscal third quarter, which ended Oct. 3. The company said at the time that 80% of its sales go to the data center market.

“There is no question that AI is reshaping hard drive demand by elevating the economic value of data and data storage,” CEO Dave Mosley said on a call with analysts.

Seagate is unlikely to have extra hard drives in inventory, Bank of America analysts wrote in November, as any additional shipments will get claimed quickly. The analysts also noted that customers are signing build-to-order contracts with firm volume and price commitments.

“The extra units are typically purchased by hyperscale or other mass capacity customers,” wrote the analysts, who recommend buying the stock.

Seagate’s trajectory looks similar to Western Digital. Analysts expect 21% revenue growth this fiscal year, followed by increases of about 15% and 6% in the next two years, according to LSEG.

Celestica, founded in 1994 as an IBM subsidiary, makes switches that connect networks together and manage the data and traffic flowing through them.

The stock is up more than 230% this year.

The company sells many of its switches to the biggest AI buyers. Sales rose 28% in the third quarter to $3.19 billion. Analysts expect revenue growth to climb from 26% this year to 33% in 2026 and 34% in 2027, according to LSEG.

Celestica CEO Robert Mionis said on an earnings call in October that a hyperscaler recently approached the company to build parts to connect liquid-cooled rack-scale computers for AI, and said mass production is scheduled to begin next year.

One boon for Celestica is surging demand for custom chips called ASICs, which are less flexible than GPUs but can be cheaper to operate for specific AI applications.

“Our largest and fastest-growing market presence is within AI data centers, supporting high-performance networking and custom ASIC AI/ML compute platforms,” Mionis said.

Analysts at Goldman Sachs wrote in a note Friday that Celestica supplies parts for Google’s ASIC.

“The company should benefit in 2026 from being the leading provider of Google TPU rack level solutions,” the analysts wrote.

WATCH: Expect a drive towards efficiencies in ai in 2026