Nvidia's AI investment portfolio swells
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Maura Losch/Axios
Nvidia has been on an investing spree the last two years, expanding the AI economy — a boom that helped make the company the most valuable in the world.
Why it matters: If that sounds a bit circular, that's what skeptics have said too, questioning whether the huge amounts of money from Nvidia and others going into AI infrastructure can ever be profitable.
Driving the news: Nvidia said on Monday that it would take a $2 billion stake in the shares of chip-software designer Synopsys as part of a broader partnership.
- Shares of Synopsys were up 3.6% in early trading.
- Last month, Nvidia joined with Microsoft to announce a $15 billion investment ($10 billion of that from Nvidia) in AI developer Anthropic, one of many deals the company has struck of late.
The big picture: By funneling money into firms whose growth drives chip demand, Nvidia has become a kind of Cornelius Vanderbilt or J.P. Morgan-style backer of the new industrial build-out.
- In 2024 and so far this year, Nvidia and its NVentures unit have made 117 investments totaling $62.24 billion, per PitchBook.
- That's a steep ramp up from the previous two years, when Nvidia made 46 such investments totaling $6.09 billion.
What they're saying: Asked about Nvidia's investments on the most recent earnings conference call, CEO Jensen Huang said, "this is really, really important work that we do."
- "All of the investments that we've done so far, all of it, period, is associated with expanding the reach of CUDA [Nvidia's parallel computing platform and programming model] expanding the ecosystem," he said, according to an AlphaSense transcript of the call.
- Nvidia did not respond to Axios' requests for comment on the PitchBook data and its investment strategy.
Yes, but: There is a logic to Nvidia's equity investments that can get lost amid eye-popping numbers, like the $100 billion partnership with OpenAI that was announced in September.
- UBS wrote in October that Nvidia's arrangement with OpenAI was "performance-based," unlike the speculative commitments common during the dot-com bubble.
- Its investment would be contingent on OpenAI putting into operation at least 10 gigawatts of AI data centers using Nvidia systems. That's a colossal effort and will take time — the first gigawatt is expected in the second half of next year.
What we're watching: Investors will have their eye on where Nvidia will invest next, and whether its strategy will change now that Google, and its in-house AI chips, have emerged as a competitor.
