Nvidia to Spend $26B on Open-Weight AI Models

In a strategic move that could reshape the AI landscape, NVIDIA is planning a colossal $26 billion investment to build its own suite of open-weight AI models, according to recent filings. This massive commitment signals the chipmaker's ambition to expand its dominance from hardware and infrastructure directly into the foundational model layer. Open-weight models, which share their architecture and trained parameters, stand in contrast to the closed models of companies like OpenAI and Anthropic. This investment could quickly position NVIDIA as a formidable competitor in the model space, offering high-performance alternatives that are optimized to run efficiently on its own industry-leading hardware. The move represents a vertical integration play of unprecedented scale. By controlling both the silicon and the core software that runs on it, NVIDIA could offer unparalleled performance and efficiency packages. It also provides a strategic hedge, ensuring that the demand for its chips is