Nvidia’s stock is still the market’s favorite AI barometer, but the latest earnings picture shows the company’s fastest-growing engine is not the GPU business everyone watches. It is networking, a less glamorous but increasingly central part of the AI buildout that helped drive total revenue to $81.6 billion in Nvidia’s first quarter of fiscal 2027, up 85% from a year earlier. The market has been rewarded for treating Nvidia as more than a chipmaker, and Wednesday’s report showed why.
What Wall Street expected before the print was already striking: networking revenue was seen approaching $17 billion in the quarter, up from about $3 billion per quarter two years ago. The actual result came in at $14.8 billion, according to the evidence pack, still a massive business and up 199% year over year. That compares with compute revenue of $60.4 billion, which rose 77% from a year earlier. In other words, the bigger business is still compute, but the faster mover is the plumbing around it.
That shift matters because the AI race is no longer only about building better chips. Alphabet, Amazon, Microsoft, Meta Platforms and other hyperscalers are all spending heavily on custom processors or commissioned silicon designed to handle some workloads without Nvidia GPUs. But those chips do not operate in isolation. They have to communicate at speed across massive data centers packed with thousands of accelerators, and that is where networking has become a critical bottleneck.
Jessica Inskip, director of investor research at StockBrokers.com, framed Nvidia’s strategy on Yahoo Finance’s Opening Bid as “investing in the bottlenecks that are the AI infrastructure.” That is the clearest lens for reading Nvidia’s push beyond the chip itself. The company is no longer just selling a processor. It is selling the systems that keep those processors from sitting idle.
Inside Nvidia’s expanding networking stack are Spectrum-X products, which connect servers over high-speed Ethernet, plus InfiniBand and NVLink, which move large amounts of data between processors quickly. Nvidia has also pushed NVLink Fusion, which lets customers plug their own custom processors into Nvidia’s broader system of high-speed connections, networking, power and cooling. The practical idea is simple: make the entire rack work like one giant computer instead of a pile of separate parts.
That broader system approach changes the competitive equation. If a customer buys a rival chip for part of the workload, Nvidia may still have a seat in the data center through networking and interconnects. That does not make the chip business less important. It does mean the company has a second growth track riding alongside it, one that is directly tied to the same AI spending boom that has powered the entire group.
The market has already started to price Nvidia as something broader than a traditional semiconductor name. Since the semiconductor trade peaked around June 22, Nvidia stock has gone essentially nowhere, while the iShares Semiconductor ETF has fallen roughly 20%. Yahoo Finance noted that Nvidia largely avoided the chip bear market that swallowed most of the group. The networking story does not explain all of that resilience, but it helps explain why Nvidia has kept its footing while much of the sector cracked.
That relative strength showed up again in the stock’s recent move. Nvidia rose 1.3% in regular trading to close at $223.47 on May 20, 2026, then fell about 0.6% in after-hours trading to around $222, according to the evidence pack. That kind of reaction fits a stock where expectations are already high and investors are scrutinizing every piece of the AI supply chain, not just the headline GPU numbers.
The scale of networking’s rise is what makes it stand out. In the prior quarter, Nvidia’s networking operation generated nearly $15 billion, up from roughly $3 billion per quarter two years earlier. Networking also grew nearly 200% year over year in that prior quarter, far faster than compute’s 77% gain. The latest reported figures confirm that networking has become one of the company’s fastest-growing pieces, even if growth is starting to slow from those extreme levels.
The quarter also showed how large the broader Data Center business has become. Data Center revenue reached a record $75.2 billion, up 92% year over year. Compute still dwarfs networking inside that segment, but networking’s pace shows the company’s revenue mix is evolving. At $14.8 billion, networking represented roughly one-fifth of Data Center revenue in the reported quarter, giving Nvidia another way to capture spending across AI infrastructure even when customers diversify their chip supply.
That is what makes the story bigger than one product line. Nvidia is not just competing with rival processors. It is trying to own more of the architecture around them. Inskip described the broader ambition more simply, saying Nvidia is trying to become a “one-stop shop for everything” in AI.
The next test is not whether Nvidia can sell GPUs. It is whether investors believe the company can keep expanding into the layer that connects, cools and coordinates the chips others are racing to build. Wall Street will keep watching Nvidia’s networking growth, custom silicon strategy and how much of the AI system it can capture beyond the GPU. The fact that analysts had looked for networking to approach $17 billion before the report shows how much attention that side of the business is now attracting.
Nvidia also announced an $80 billion share repurchase authorization and raised its quarterly dividend, a reminder that management is returning cash even as it spends aggressively to stay ahead in AI infrastructure. The company reported total revenue of $81.6 billion in the quarter, and it followed with guidance for the next quarter that implies a midpoint of $91 billion. That gives investors a fresh benchmark, but the more important signal may be structural rather than cyclical.
The story now is that Nvidia’s fastest-growing business is no longer the part that made it famous. The market still calls it a chip company, but the numbers say it is becoming something wider: a networked AI infrastructure platform with a faster-growing business beside the GPUs everyone talks about.
