Nvidia Will Spend $26 Billion to Build Open-Weight AI Models, Filings Show (www.wired.com)

Nvidia has announced an ambitious plan to invest $26 billion over the next five years into developing open-source AI models, a significant shift that positions the tech giant as a potential leader in open AI research, akin to OpenAI and DeepSeek. This move aims to solidify Nvidia’s dominance as the premier chip manufacturer in the AI sector, as these models will be specifically optimized for its hardware. The recent launch of the Nemotron 3 Super model further underscores Nvidia's commitment; with 128 billion parameters, it claims superior performance over existing models, including OpenAI’s GPT-OSS, and has achieved top scores on various AI benchmarks. This strategic investment in open-source development could reshape the competitive landscape, particularly against the backdrop of freely available Chinese AI models. Nvidia's open models will not only foster innovation by allowing startups and researchers to build upon its technology, but they will also enhance the capabilities of Nvidia’s own hardware and supercomputing infrastructures. As the race for AI dominance intensifies globally, Nvidia's dual focus on accessibility through open models and performance optimization could prove critical for maintaining its edge, while providing a strong alternative to Chinese developments in the AI space.