Nvidia bulks up open source offerings with an acquisition and new open AI models

Nvidia is making significant strides in the realm of open source AI through two key initiatives: a strategic acquisition and the unveiling of new AI models. The tech powerhouse announced on Monday that it has acquired SchedMD, the foremost developer behind the widely-used open source workload management system, Slurm. SchedMD, founded in 2010 by lead developers Morris Jette and Danny Auble, will continue to manage Slurm as an open source, vendor-neutral software designed specifically for high-performance computing and AI applications. The financial details of this acquisition have not been disclosed, and Nvidia has opted not to provide further comments beyond its official blog post. Nvidia has a long-standing collaboration with SchedMD, spanning over a decade, and emphasizes that this technology serves as critical infrastructure for generative AI. The company intends to enhance its investment in this technology, aiming to broaden accessibility across various systems. In addition to the acquisition, Nvidia introduced a new lineup of open AI models on the same day. Dubbed the Nvidia Nemotron 3, this family of models is touted as the most efficient collection of open models designed for developing precise AI agents. The lineup features several variants: the Nemotron 3 Nano for specialized tasks, the Nemotron 3 Super for applications involving multiple AI agents, and the Nemotron 3 Ultra, which is tailored for more complex challenges. Jensen Huang, Nvidia's founder and CEO, highlighted the company's commitment to open innovation, stating, "With Nemotron, we’re transforming advanced AI into an open platform that gives developers the transparency and efficiency they need to build agentic systems at scale." Nvidia’s recent activities underscore its dedication to enhancing its open source and open AI capabilities. Just last week, the company announced the launch of Alpamayo-R1, a new open reasoning vision language model aimed at advancing autonomous driving research. Additionally, Nvidia has expanded its resources with more workflows and guides for its Cosmos world models, which are available under a permissive open source license, facilitating developers in harnessing these models for physical AI applications. This strategic momentum reflects Nvidia’s vision that physical AI will be the next significant frontier for its GPU technologies, positioning the company as a leading provider for robotics and self-driving vehicle firms seeking advanced AI and software solutions.