- Nvidia has made four acquisitions so far in 2024, matching its 2020 total.
- Regulatory scrutiny is delaying the closure of some of Nvidia's deals, including Run:ai.
- Experts say AI mergers and acquisitions will slow, but not stop as regulators scrutinize the space.
Nvidia is entering the summer of what could be its most acquisitive year yet.
According to public announcements, the firm has made four acquisition deals so far in 2024, exactly even with Nvidia's 2020 deal total. But some of this year's deals have yet to close, in part due to regulatory scrutiny.
Nvidia, along with Microsoft and OpenAI, is the subject of a Department of Justice and Federal Trade Commission investigation into possible antitrust violations.
Nvidia declined to discuss its acquisition strategy for this report, citing a quiet period before it announces earnings on August 28. It also declined to confirm its acquisition count for this year or if it is indeed tied for most deals with 2020.
That was a remarkable year, and not only because of the wrenching global pandemic panic followed by a surprising economic rebound. In April of 2020, Nvidia acquired Mellanox for $6.9 billion. The Israeli networking company has been touted as key to the AI data center strategy as we know it. And in September 2020, Nvidia announced an ill-fated $40 billion deal to acquire chip designer ARM from Softbank. The parties terminated the deal in February 2022 citing regulator interference.
Nvidia's latest deals don't necessarily rise to the profile of those, but they're notable all the same since the company has seemingly limitless resources and a firm spot as the foundation of the AI ecosystem.
Meet the deals
In April Nvidia agreed to acquire Israeli firm Run:ai, but the deal has not yet closed and is receiving scrutiny from the Department of Justice, according to Politico. Run:ai's technology makes GPUs run compute more efficiently, allowing more work to be done on fewer chips.
In May, Nvidia agreed to acquire Deci AI, another Israeli startup that makes tools for developers to build AI models, for $300 million.
Months later in July, investors in California firm Shoreline confirmed that Nvidia would acquire the company, which uses automation to reduce and address service-disrupting incidents in cloud computing.
Later in July, Nvidia acquired Brev.dev, run by a small team out of San Francisco that seeks to lower the barrier to entry for AI development. The two companies had been collaborating for roughly 11 months, according to founder Nader Khalil speaking in a video posted to LinkedIn. The value of this acquisition remains undisclosed and may fall far below the usual threshold for regulator scrutiny.
Though the company isn't keen to discuss its strategy publicly, Nvidia executives recently gave some indication of what it looks for.
"Small and medium M&A will always be an area that we research. We take a strong look at these types of companies that we think could help bolt on," said CFO Colette Kress at a Citi Global technology conference in 2023. At another investor conference around the same time, Kress said it's unlikely the company would find another acquisition as consequential as Mellanox, but would still look for small companies with additive technologies.
Slower and quieter but still going
AI industry watchers and investors told Business Insider that acquisitions are slowing down among the big technology players who are under the microscope from the DOJ and FTC. But small deals are still going through and even some larger ones.
In June, Amazon hired the majority of the team from Adept, a company with a foundation model aiming to rival OpenAI, and licensed the company's technology in a move commonly called a "reverse acquihire."
In July Nvidia competitor AMD announced an agreement to acquire Silo AI, a Finnish AI startup, for $655 million.
And this quiet trickle of deals in various forms is unlikely to stop since the market for AI talent is so tight.
"It's $200,000 to $800,000 for an AI engineer right now," Paul Baier, CEO of GAI Insights, said. And that's if you can find one.
Acquihires are a common solution in such conditions. "When you do an acquihire for a failed AI startup, it's $4 million per engineer. That's how scarce this talent is," Baier continued.
High valuations for startups that raised funds amid the initial AI frenzy, plus a severe talent shortage, equals, more acquihire deals, Umesh Pavdal, managing director of Thomvest, said, even with the more intense regulatory attention.
Got a tip or an insight to share? Contact senior reporter Emma Cosgrove at [email protected] or use the secure messaging app Signal: 443-333-9088.