By Dan Gallagher

Arm Holdings picked a perfect time jump into the AI market with its own chips. Which is a good thing, as its stock price already assumes the effort will go perfectly.

The British chip designer has long been in the business of farming out designs. That has been a massive success, as nearly every mobile device plus a growing number of personal computers and servers use Arm-based processors that are made by well-known companies such as Nvidia, Apple, Samsung and Qualcomm. But Arm now wants a bigger piece of the action, evidenced by the first launch of its own chip at a high-profile event in San Francisco this week.

That chip, which is designed for artificial-intelligence servers, already has investors cheering. Arm's share price jumped 16% on Wednesday and has surged 50% since the company's fiscal third-quarter report on Feb. 4, where it first teased the event. But those gains also add to Arm's already pricey valuation; the stock now trades at 81 times projected earnings.

That is one of the highest multiples in semiconductors -- and nearly four times that of AI leader Nvidia. In other words, the stock price bakes in a lot of early success for Arm's first chip. It also assumes the company can pull off a major business-model shift without endangering the relationships that have made it so successful. Nvidia, a major customer for Arm's basic designs, is stepping up its own CPU efforts in a big way. Other major Arm customers include tech giants like Amazon, Microsoft and Google which design Arm-based chips in house for their own data centers.

Arm also may seem late to the game, given that the AI spending boom is now in its third year. But the timing is actually shrewd. AI computing needs are shifting to the types of central processing chips -- or CPUs -- that are needed to power inferencing, where AI models produce answers to user queries, rather than model training.

The adoption of AI agents is expected to push the need for inference even further. But AI data centers still demand a lot of energy, so Arm's roots in designing CPUs that have to run on batteries is a big advantage. Mark Lipacis of Evercore ISI said Arm's designs make it perfectly matched for a market needing more power-efficient chips to run agents. "Agentic to Arm is like AI to Nvidia," Lipacis wrote in a report Wednesday.

Those advantages have helped Arm land some important early customers. Meta Platforms and OpenAI will be using Arm's new chip. Financial terms weren't disclosed, though Arm described the relationship with Meta as "multi-generation." Notably, Arm projected annual revenue from its new chip program reaching $15 billion for the fiscal year ending in March of 2031. That is 50% higher than what the company expects for its licensing business in the same period.

Such growth would make Arm's overall business more than five times as large as it is now. But it won't come without a cost. Licensing intellectual property is a super profitable business that allows Arm to command gross profit margins of 97% now. Making its own chips requires use of manufacturing partners and component suppliers that will raise the company's costs. Arm told analysts on Tuesday that it expects the chip side of the business to garner gross margins of at least 50%.

Wall Street sees the effort as worth the cost, as selling chips directly allows Arm to realize much more revenue for its designs. At least 10 analysts have raised their price targets on the stock following Tuesday's event, according to FactSet data. And few see much risk in Arm hurting its relationships with cloud-computing giants that license its designs, as Arm's new chip is mainly targeting the wide swath of corporate customers who aren't developing their own processors.

As for Nvidia, it doesn't seem too worried. Chief Executive Officer Jensen Huang even appeared in a video at the Arm event, congratulating the company on the launch. For his part, Arm CEO Rene Haas thinks there is plenty of business to go around in CPU chips for AI. He told analysts on Tuesday that "there is a very, very large market here where multiple players can play." That better be the case, because at 80 times forward earnings, Arm can't afford to play around.

Write to Dan Gallagher at dan.gallagher@wsj.com

(END) Dow Jones Newswires

March 26, 2026 05:30 ET (09:30 GMT)

Copyright (c) 2026 Dow Jones & Company, Inc.