Microsoft and Nvidia Are the Odd $4 Trillion Power Couple

Microsoft and Nvidia Are the Odd $4 Trillion Power Couple

Nvidia and Microsoft could soon become $4 trillion companies, forming the most exclusive club in the stock market.

PREMIUM Nvidia CEO Jensen Huang and his counterpart Satya Nadella at Microsoft.

Enthusiasm over artificial intelligence is what got both of them there. But for Microsoft, the story is more complicated—and the payoff fuzzier.

A much bigger company than Nvidia in terms of annual revenue, Microsoft also sits in a different spot in the AI value chain. Nvidia is booming because any company that wants to work with AI buy its chips first">has to buy its chips first. Microsoft’s boom will depend on the willingness of a high number of the customers it depends on—both businesses and consumers—to pay a premium for AI services.

For that to happen, AI will have to become a more transformative corporate tool fully woven through everyday life—much like Microsoft’s Windows operating system or Word office software became decades ago.

Many think that’s more a question of “when” than “if.” But the “when” still matters after Microsoft’s market value has grown by a trillion dollars in less than three months. And, at a market cap of $4 trillion, Microsoft’s stock price would command the highest multiple against projected earnings that it has sported in more than 20 years.

That leaves no room for error—and relatively little room even for speed bumps.

Microsoft has had some of those of late. Its early partnership with OpenAI gave the company a front seat in the AI rocket ship, allowing it to infuse the technology behind ChatGPT into its own products.

But that relationship has hit a rough patch. OpenAI wants to change its unusual corporate structure and become a regular for-profit company, part of a long-running effort to disentangle itself from a benefactor that is also a competitor.

Worryingly for Microsoft, OpenAI has the right to limit access to its future technology when it reaches “artificial general intelligence”—a nebulous threshold that could still deprive Microsoft of a crucial driver of its AI strategy.

Microsoft has also reportedly had trouble developing its own line of in-house AI chips that could lessen its dependence on Nvidia. Adding to the internal drama, the company confirmed Wednesday that it plans to lay off 9,000 more workers, on top of the 6,000 roles it eliminated in May.

Other tech giants have made cuts with the aim of boosting efficiency and investing more in AI. But Microsoft might still have a lot more ground to cover here.

The company commands the lowest annual revenue per employee among its big-tech peers save for Amazon, according to data from S&P Global Market Intelligence. Microsoft would need to cut its last-reported head count by 84,000 positions to simply catch up to Google-parent Alphabet on that metric.

But taking costs out will only take Microsoft so far. The company still needs to make AI more than a bit player on its income statement.

Analysts estimate that AI services within Microsoft’s Azure cloud-computing arm generated revenue of $11.5 billion for the just-ended fiscal year, according to consensus estimates from Visible Alpha. That’s more than double the prior year but still only about 4% of the company’s total annual sales.

The limited contribution of AI so far does give Microsoft some downside protection if that business doesn’t pan out—or even grow at the rate its enthusiasts envision.

That’s a notable contrast to the more binary question faced by Nvidia. The chip maker’s annual sales have surged more than 10-fold over the last three years and are still expected to average 32% growth annually over the next three years.

But that could vanish quickly if AI demand doesn’t materialize for its biggest customers—or if other technological breakthroughs render its chips less vital. Nvidia lost 20% of its market value in just a week’s time in January after claims by Chinese AI startup DeepSeek raised the possibility that advanced AI models could be produced without having to spend billions on Nvidia’s chips.

Announced capital spending plans by Nvidia’s largest customers strongly suggest that’s unlikely to happen soon. Microsoft, though, still needs AI to actually become a ubiquitous and transformative corporate tool.

The company behind Windows and Office is large and well-heeled enough to help make that happen. But it isn’t clear whether it will take place as quickly as Microsoft needs to justify an ever-higher valuation.

As Fed Chair Jerome Powell put it in congressional testimony last week, it usually takes more time than people expect for world-changing technologies to be implemented in ways that shift the economy.

At its current size and valuation, timing is going to be everything for Microsoft.

Write to Dan Gallagher at dan.gallagher@wsj.com and Asa Fitch at asa.fitch@wsj.com