Nvidia-Groq deal keeps fiction of competition alive, analyst says

It’s been two days since the news broke that Nvidia was spending $20 billion to acquire top talent from Groq in what the chip startup called a “non-exclusive licensing agreement.”

Nvidia, the world’s most valuable company, has not issued a press release or regulatory filing and, according to a spokesperson, is only confirming the contents of Groq’s 90 words. blog post released Wednesday after the close of holiday-shortened trading.

“They’re so big now that they can do a $20 billion deal on Christmas Eve without a press release and without anyone batting an eye,” Stacy Rasgon, an analyst at Bernstein, said in an interview with CNBC on Friday.Screamed in the street“.

Although neither company has confirmed the price, CNBC learned Wednesday from Groq lead investor Alex Davis that Nvidia has agreed to buy assets of Groq, a designer of high-performance artificial intelligence accelerator chips, for $20 billion in cash. Davis’ company, Disruptive, invested more than half a billion dollars in Groq and led the startup’s development. last round of the table in September at a valuation of $6.9 billion.

Groq founder and CEO Jonathan Ross, along with Sunny Madra, the company’s president, and other senior executives “will join Nvidia to help advance and scale licensed technology,” the startup said in the post, adding that it would continue as an “independent company,” led by CFO Simon Edwards.

As an acquisition, Groq would be by far Nvidia’s largest in its 32-year history. Its biggest previous purchase came in 2019, when Nvidia bought an Israeli chip designer. Mellanox for almost 7 billion dollars.

But Nvidia rather follows a game book used by other tech giants over the past two years, spending billions of dollars to hire top AI talent and gain access to key technologies through licensing deals.

This is a strategy that has been used by Meta, Google, Microsoft And Amazon. Nvidia itself has used this tactic before, shelling out more than $900 million in September to hire Enfabrica CEO Rochan Sankar and other employees from the AI ​​hardware startup, and to license the company’s technology to CNBC. reported at the time.

By avoiding traditional acquisitions, tech companies have been able to bypass some level of antitrust scrutiny and quickly make deals to attract the people they want most.

“Antitrust appears to be the primary risk here, although structuring the deal as a non-exclusive license could maintain the fiction of competition,” Rasgon wrote in a note to clients Thursday. His firm recommends buying Nvidia stock and has a $275 price target for the stock.

Nvidia shares rose about 2% on Friday to $192.40. The stock is up 43% this year and has risen 13-fold since late 2022, when generative AI began to take off following the launch of OpenAI ChatGPT.

Nvidia has used its expanding cash pile to invest capital into the AI ​​ecosystem, including through recent investments in OpenAI and Intel. At the end of October, Nvidia had $60.6 billion in species and short-term investments, compared to $13.3 billion at the start of 2023.

Groq was based in 2016 by a group of former engineers, including Ross. He was one of the creators of Google’s tensor processing units, or TPUs, the search giant’s technology. custom chips which are used by some companies as an alternative to Nvidia’s graphics processing units or GPUs.

The startup’s specialty is on the inference side of the market, which refers to using AI to make decisions based on new information. Nvidia dominates the training industry, which involves teaching AI models to learn from patterns derived from large amounts of data.

Cantor analysts said in a report Friday that Nvidia was “playing both offense and defense” by seizing Groq’s assets, preventing them from falling into the hands of a competitor.

“We believe this acquisition only strengthens Nvidia’s comprehensive systems stack and overall leadership in the AI ​​market (and only widens its competitive moat),” wrote the analysts, who maintained their Buy rating and $300 price target.

Analysts at BofA Securities also maintained their buy recommendation and target of $275 following the announcement. In a note released Friday, they called the deal “surprising, expensive but strategic,” and said it shows Nvidia recognizes that “while the GPU dominates AI training, the rapid transition to inference may require more specialized chips.”

Analysts said key questions remain, such as who will own the intellectual property for Groq’s language processing unit, whether it can be licensed to Nvidia’s competitors and whether what’s left of Groq – its nascent cloud business – could potentially “undercut NVDA’s LPU-based service with lower prices.”

Nvidia is not commenting on any of these details at this time. The first opportunity analysts and investors will likely have to hear about the company will be on January 5, when CEO Jensen Huang is expected to visit the company. speak at CES in Las Vegas.

—David Faber of CNBC contributed to this report

WATCH: From Nvidia GPUs to Google TPUs – Breaking Down All AI Chips