(Bloomberg) -- Nvidia Corp. is in early discussions with the Korean AI chip designer Rebellions about possible collaborations, including a technical partnership, an investment or perhaps even an acquisition, according to people familiar with the matter.
Most Read from Bloomberg
Trump Sends First Haiti Deportation Flight After Ending TPS for 350,000 - China Sentences Evergrande's Hui to Life for 'Heinous' Crime - Bessent Deploys Debt Buybacks in Sign of Concern Over Yield Rise
Nvidia Chief Executive Officer Jensen Huang met with Rebellions co-founder and CEO Sunghyun Park and discussed a potential tie-up at the US company's headquarters in Santa Clara, California this week, said the people, who asked not to be identified as the information is not public. Huang has led a far-reaching investment spree in recent years, as Nvidia works to maintain its central role in the artificial intelligence era.
Those deals have included high-profile startups. In late 2025, Nvidia struck an unconventional deal with US AI firm Groq, securing a nonexclusive license while absorbing most of its engineering talent. Nvidia said it wasn't an acquisition because Groq continues to operate separately and run its cloud business independently.
Bundang, South Korea-based Rebellions has raised roughly $850 million from investors including SK Hynix Inc., Samsung Ventures and Arm Holdings Plc, and was most recently valued at around $2.3 billion. It has also secured a direct investment from the Korean government.
Deliberations between Nvidia and six-year-old Rebellions are preliminary and may not lead to a transaction, the people said. Nvidia did not respond to requests for comment. A representative for Rebellions declined to comment.
Founded in the midst of the pandemic in 2020, Rebellions is among a group of semiconductor startups designing and deploying their own chips to rival market leader Nvidia. The company specializes in neural processing units for data centers, optimized for AI inference work. Less intensive in processing requirements than AI training, inference is the larger-in-volume workload required to operate AI services at scale.
Nvidia holds such a dominant share of the semiconductors used to train AI models like those from OpenAI and Anthropic PBC that it has to be wary of antitrust scrutiny. Large acquisitions in the tech industry typically need clearance from authorities such as the US Department of Justice.