The author is a Reuters Breakingviews columnist. The opinions expressed are his own.

By Robert Cyran

NEW YORK, May 20 (Reuters Breakingviews) - Jensen Huang has compared being Nvidia’s NVDA.O boss to being a conductor. His job at the $5 trillion chipmaker is to ensure that the bewilderingly large number of executives that directly report to him work together to produce a harmonious silicon sonata. Something similar could be said about the company’s role in ensuring lots of data centers, full of its gear, get built.

Nvidia conducts this breakneck, AI-fueled industrial shift with its growing financial heft. Just look at its results for the first quarter, unveiled on Wednesday. The company posted revenue of $81.6 billion, an increase of 85% from the same period last year. Cash from operations increased 147% to $53.5 billion. Nvidia also raised its dividend, and authorized an additional $80 billion of share repurchases.

Making this happen is no solo act. Producing Nvidia’s leading AI chips requires pulling together a vast network of suppliers to provide parts like memory, as well as to assemble, package and test servers and networking gear. Huang must cajole, wheedle, and inspire everyone along that chain: Nvidia’s reputation for showing suppliers a good time helps, as does the size of the opportunity. As of January 25, the company had $95 billion of outstanding commitments, mostly due this fiscal year, and additional implicit guarantees.

Nvidia can also use its substantial resources more forcefully. About a third of its cash from operations last year went towards investments in other firms. It paid $13 billion on closing for a non-exclusive licensing deal with upstart chip designer Groq, alongside $17.5 billion to buy stakes in other private companies.

Stakes in public companies, meanwhile, ballooned to over $17 billion. While a giddily rising market helped the likes of its position in Intel INTC.O, the growth of this figure mostly represents new investments.

The optimistic case is that this helps further coordinate around AI bottlenecks. Need to speed up data transfer? Invest $3 billion in glassmaker Corning and help fund new factories in hopes of optical connectivity. Model makers like Anthropic or OpenAI need infrastructure? Promise to take a stake in return for what is essentially vendor financing.

It’s difficult for anyone to oversee dozens of people, let alone a burgeoning industry. These entanglements also make Nvidia more difficult to understand. In the case of an AI demand downturn, direct revenue would fall, obviously. Less clear is what happens to those swelling supplier commitments. The value of its investments, especially in unlisted firms, might crater. A similar story played out among telecommunications suppliers 20 years ago. Nvidia is much stronger than they ever were. Add too many members to the orchestra, though, and you begin to risk discord.

Follow Robert Cyran on Bluesky.

CONTEXT NEWS

Nvidia said on May 20 that revenue for the quarter ending April 26 was $81.6 billion, an increase of 85% from the same period last year. Cash from operations increased 147% to $53.5 billion.

The chipmaker also raised its quarterly dividend from 1 cent per share to 25 cents per share, and announced an additional $80 billion share repurchase authorization.