Nvidia is quietly building a multi-billion dollar behemoth to compete in its chips business

Nvidia CEO Jensen Huang was ahead of the market when he pushed the company to start playing with it. building AI chips in 2010more than a decade before the current buzz around AI. A similar move in 2020 – doubling down on data center networks and acquisitions – has resulted in a highly profitable and fast-growing division, but less popular.

In just a few years, Nvidia’s Internet business, designed to connect data centers, has grown to become the second largest company by revenue behind the computing industry. Last quarter, it generated $11 billion in revenue, a year-on-year increase of 267%, and brought in more than $31 billion in revenue for the entire year, according to Nvidia. recent findings.

Driven by the growth of AI processing, these components include technologies such as NVLink, which enables communication between GPUs in a data center rack, Nvidia InfiniBand Switches, a network computing platform, Spectrum-Xan ethernet platform for AI communications, and integrated Optics switches, among others.

Together, Nvidia’s online business combines all the technologies needed to build an “AI factory,” a data center designed to train AI models.

Kevin Cook, chief economist at Zacks Investment Research, told TechCrunch that Nvidia’s internet business is one of the company’s most exciting new divisions. “(Nvidia’s Internet business) is said to be $11 billion a quarter; this figure is greater than Cisco’s Internet business, almost the same as the annual estimate,’ Cook said, adding that it does in one quarter what Cisco’s business does in a year.

And yet – the business sector does not attract the same attention as the corporate business, which is much larger. It doesn’t get as much excitement as the company’s gaming business, it’s the bread and butter startup business, which is three times smaller.

The origins of Nvidia’s Internet business come from Mellanox, an Internet networking company founded in Israel in 1999. Nvidia earned in 2020 for $ 7 billion.

Kevin Deierling is senior vice president of Internet at Nvidia and joined the company through the Mellanox acquisition. Deierling told TechCrunch that people not knowing about Nvidia’s internet business could be the reason for its poor sales performance — but he doesn’t like that answer.

“People think of social media as, ‘I have a printer, and I need to connect to it,'” Deierling said. “Jensen said this the first day he found us, he said data centers are a new area of ​​computing. Networking is more than just moving bits of data between compute nodes, it’s actually the foundation.”

Although Deierling said he didn’t understand why Huang bought the company when he did – he gets it now. Having an online business alongside its GPU business allows the company to sell smaller chips and technologies that it specializes in.

“When Jensen bought Mellanox in 2020, he saw it as the missing piece to make GPUs a complete package,” said Cook, the Zack analyst.

Deierling added that he thinks that part of the success of Nvidia networks is that they only sell technology as a complete solution, as opposed to individual components, and they do not sell technology alone, but through partners.

“I can’t think of any other company that has (all) of the skills that we have,” Deierling said. “We’re very different. We make all the integrated, well-integrated products, and then we go to market through our partners.”

Nvidia has just announced a new execution updates on its network at Huang’s big event on March 16 of the company’s annual event Nvidia GTC technical conference. The company launched the Nvidia Rubin platform, which includes six new chips to power the “AI supercomputer.” Nvidia also announced a new Nvidia Inference Context Memory Storage platform and Nvidia Spectrum-X Ethernet Photonics smart switches, among other things.

“It’s no longer a peripheral to connect to a printer, just another slow I/O device,” Deierling said of networking. “It’s very important to the computer. In the old days, we had what’s called the backend inside the computer. Today, the network is the back line of the AI ​​factory, and it’s very important.”