Prolific Cisco inventor JP Vasseur is joining Nvidia shortly after the AI computing giant’s CFO, Colette Kress, said that the company’s Spectrum-X line of Ethernet networking products for data centers is ‘well on track to begin a multibillion-dollar product line within a year.’

Nvidia has hired a 25-year Cisco Systems engineering veteran, once credited as the switching giant’s most prolific inventor, who will lead development of AI and networking architecture at the AI computing giant.

JP Vasseur, a recently departed Cisco fellow who was most recently vice president of engineering of machine learning and AI for networking, announced in a LinkedIn post on Wednesday that he joined Nvidia this month as a senior distinguished engineer and chief architect of AI and networking.

[Related: Analysis: How Nvidia Surpassed Intel In Annual Revenue And Won The AI Crown]

“With over 32 years of experience in networking, including the past 13 years focused on AI and networking products, I am invigorated by the immense opportunities ahead to drive innovation in these areas,” he wrote in a LinkedIn post.

An Nvidia spokesperson declined to comment. Vasseur and Cisco did not respond to requests for comment.

Nvidia Makes Multibillion-Dollar Ethernet Sales Push

Vasseur’s announcement comes a little more than a month after Nvidia CFO Colette Kress said that the company’s Spectrum-X line of Ethernet networking products for data centers is “well on track to begin a multibillion-dollar product line within a year.”

Kress made the comment during Nvidia’s second-quarter earnings report in late August and also said that the company’s Ethernet for AI revenue, which includes Spectrum-X products, doubled sequentially from the previous quarter.

“Spectrum-X has broad market support from OEM and ODM partners and is being adopted by [cloud service providers], GPU cloud providers and enterprise, including [Elon Musk’s] X.ai, to connect the largest GPU compute cluster in the world,” she said.

Claiming that Spectrum-X “supercharges Ethernet for AI processing and delivers 1.6 [times greater] the performance of traditional Ethernet,” Kress said Nvidia plans to launch new Spectrum-X products every year—alongside new chips—to “support demand for scaling compute clusters from tens of thousands of GPUs today to millions of GPUs in the near future.”

The decision for Nvidia to hire a Cisco engineering veteran with experience at the intersection of AI and networking should serve as another sign that the AI computing giant is making networking a major priority, a principal engineer at a solution provider told CRN.

“They want to focus on networking in a big way, which makes sense why that guy went over to them,” said the engineer, who asked that he not be named because he wasn’t authorized to speak as a representative of his company.

Nvidia started making a big push into networking when it acquired high-speed interconnect vendor Mellanox Technologies for $7 billion in 2019. Since then, Mellanox’s technologies have played a significant role in Nvidia’s strategy to become a “data center-scale computing company,” as Nvidia CEO Jensen Huang once said.

While the company sells a variety of networking products, from Ethernet and InfiniBand switches to data processing units and SmartNICs, Nvidia uses these products as essential building blocks for AI servers and data centers that are designed in house and increasingly sold to a variety of customers, from enterprises to cloud service providers.

With Nvidia recently joining the Ultra Ethernet Consortium alongside other tech giants to develop specifications for high-speed networking solutions, the principal engineer at the solution provider said he wouldn’t be surprised if Nvidia eventually expanded its networking sales push to cover a much wider portion of the data center market.

“I think they might be banking on the fact that they’ll get out of just deploying Mellanox networking and [the company’s] BasePod and SuperPod [AI supercomputer clusters] and start making that be a part of enterprise networks as they begin to go down this Ultra Ethernet Consortium-type approach,” he said.

Vasseur: A Celebrated Engineer And Top Inventor At Cisco

With a doctorate and master’s in computer systems networking and telecommunications from Télécom Paris in France and the Stevens Institute of Technology in New Jersey, respectively, Vasseur spent the past 25 years at Cisco where he was a celebrated engineer and top inventor who worked on high-priority initiatives.

A blog post by Cisco in 2019 called Vasseur the company’s “top inventor,” with 483 issued patents carrying his name and “spanning key tech areas from machine learning and AI in security in enterprises to the Internet of Things.”

“He is one of only 19 Cisco Fellows out of 30,000 engineers, constantly pushing the boundaries of technology,” Cisco said in the blog post.

In his most recent role as vice president of engineering of machine learning and AI for networking, Vasseur said he led the development of “large language model and generative AI use case specifications, system architecture design, prompt engineering, model tuning and sophisticated techniques such as Knowledge DB and [retrieval-augmented generation],” according to his LinkedIn profile.

He previously served as the head of engineering for various high-profile areas including predictive networks, IoT and advanced threat detection.

When Vasseur announced that he was leaving Cisco two weeks ago, he said, “one of the most exciting parts of my journey has been contributing to the birth and explosive growth of a new industry: the Internet, to which Cisco made significant contributions.”

Now at Nvidia, Vasseur seemed to indicate that he sees his work at the AI computing giant as another journey that will bring about great change.

“Jensen Huang’s vision for Nvidia to build the new area of accelerated computing and industry with the AI factory is incredibly inspiring, and I am thrilled to join this journey and a highly talented team!” he wrote on LinkedIn.