Summary:
Nvidia is investing heavily to strengthen the U.S. open-weight AI ecosystem and compete with powerful Chinese models such as DeepSeek and Kimi.
Under the deal, Nvidia will:
- Invest $1 billion in Poolside at a $12 billion pre-money valuation.
- Pay another $6 billion to license Poolside’s AI technology.
- Hire more than 100 Poolside employees, mainly engineers, to work on Nvidia’s Nemotron open-weight models.
Poolside’s founders and senior executives will remain independent and pursue unspecified research projects. Its technology, including the popular Laguna S model, will be incorporated into Nvidia’s efforts to develop larger and more capable versions of Nemotron—potentially models with more than a trillion parameters.
The agreement reflects CEO Jensen Huang’s strategy of supporting both sides of the AI market. Nvidia supplies chips to companies such as OpenAI and Anthropic, which primarily build closed models, but it is also investing in open-weight systems that are cheaper to operate, customizable, and downloadable by users.
Nvidia believes the U.S. has fallen behind China in open AI because American labs have focused more on proprietary models. The Poolside acquisition is intended to help close that gap, though it also places Nvidia in competition with some of its biggest customers and other AI companies it supports.
For Poolside, the deal solves a major financing and computing problem. The startup failed to raise $2 billion in 2025 to secure a large Nvidia server cluster and feared running out of computing capacity. Nvidia’s capital, infrastructure, and engineering resources offered a way for Poolside’s technology to continue scaling.
Source: https://archive.is/tUJvM