Nvidia has struck a $6 billion deal with AI startup Poolside to license its technology and hire most of its engineers, according to multiple reports. The chipmaker also plans to invest $1 billion in Poolside at a pre-money valuation of $12 billion, bringing its total commitment to $7 billion. The transaction is expected to help Nvidia build open-weight AI models that can compete with offerings from China's DeepSeek and Moonshot, as reported by several outlets.
Deal Details and Strategic Intent
The Wall Street Journal reported over the weekend that Nvidia would pay $6 billion to license Poolside's technology and hire the bulk of its engineers. An additional $1 billion equity investment at a $12 billion pre-money valuation would bring the total to $7 billion, a four-fold jump from Poolside's earlier valuation of $3 billion in 2025. Over 100 Poolside employees are likely to join Nvidia, with some reports specifying 109, and will work on Nvidia's Nemotron family of open-weight AI models, according to the reports.
Poolside, founded in 2023, is known for its Laguna series of open-weight models, which are small enough to run on a single Nvidia DGX GPU. Open-weight models are generally cheaper to run and allow developers to customize them for specific needs, as noted by Silicon Republic.
Nvidia's move is seen as a response to the growing prominence of Chinese open-weight models. Data from Hugging Face, cited by Silicon Republic, showed that Alibaba's Qwen models were downloaded over 3 billion times in the past six months, far ahead of Google, OpenAI, Meta, and other rivals.
Competing with China and US Rivals
The deal comes amid concerns in the US tech industry following DeepSeek's release of a low-cost model in January 2025, which raised questions about America's lead in AI, as noted by Times Now. Nvidia's efforts to develop its own open-weight models place it in direct competition with OpenAI and Anthropic, both of which it backs, according to multiple reports.
Nvidia CEO Jensen Huang has been a vocal advocate for open-weight AI. In a recently published letter, he wrote: "Our AI leadership will be judged not by one frontier AI model, but by whether the United States builds a strong, open ecosystem that diffuses into every sector." He argued that open weights expand access to the AI economy and that relying solely on closed models is not inherently safe, as they can be breached or fail in ways that outsiders cannot detect, as reported by Silicon Republic.
Poolside has framed the agreement as supporting a future in which AI is not controlled by a small number of companies. The founders wrote that artificial general intelligence should "not be a closed technology controlled by few but one built by many out in the open," according to Times Now.
Nvidia's Growing Deal-Making and Hedging Strategy
The Poolside deal is the latest in a series of strategic moves by Nvidia to secure AI talent and technology. In 2025, Nvidia signed a licensing deal with Groq that saw much of its talent hired away by Nvidia, and it spent over $900 million to hire Enfabrica CEO Rochan Sankar and other employees while licensing Enfabrica's technology, as reported by Silicon Republic. Nvidia is also in talks with South Korean chipmaker Rebellions for a similar partnership that could result in an investment or acquisition, according to the same source.
Some analysts view the deal as a hedge against potential declines in demand for Nvidia's GPUs)Skip, with Wccftech suggesting it transforms Nvidia into a buyer of last resort for its own chips. This comes amid broader industry discussions about AI adoption timelines. OpenAI CEO Sam Altman conceded he was wrong on his aggressive AI adoption timeline, saying the economy will incorporate AI at a much slower pace than previously believed, as noted by Wccftech.
Price Hikes and Market Context
Alongside the Poolside deal, reports have circulated about Nvidia's plans to raise prices for its Grace Blackwell GPUs and Vera Rubin systems. Bloomberg reported a rise of around 15 percent early next year, while The Information pegged the increase at 17 percent, which would push the expected price of a single Vera Rubin NVL 72 rack to around $8 million, according to Wccftech. Such a price hike would add roughly $5 billion to the cost of a 1GW data center, bringing its total cost to nearly $60 billion in the US, the same outlet reported.
These developments come as China continues to close the AI gap with the United States despite restrictions on advanced AI chip supply. Bloomberg reported that Chinese models are becoming comparable in capabilities to US ones at much lower development costs, citing Kimi K3, which came close to Anthropic's flagship model's performance while its developer Moonshot spent significantly less on its creation, as reported by TASS.
Looking Ahead
Nvidia is betting that its combination of computing power, engineering talent, and Poolside's technology will help Nemotron compete with some of the world's most advanced AI models in the coming year, as noted by Times Now. The deal's non-exclusive nature and the fact that Poolside's leadership will not join Nvidia suggest a continuing independent role for the startup, according to Silicon Republic.
As Nvidia expands its footprint in the AI model space, it moves into direct competition with the very companies that rely on its chips—OpenAI and Anthropic—while also positioning itself against China's open-weight leaders. The outcome of this strategic bet will likely shape the competitive landscape of the AI industry in the years ahead.