Nvidia announced on Monday its acquisition of SchedMD, developer of the open-source workload manager Slurm, and the release of a new family of open AI models called Nvidia Nemotron 3 to expand its open-source AI offerings.
SchedMD serves as the leading developer of Slurm, a workload management system originally launched in 2002 and designed specifically for high-performance computing and artificial intelligence workloads. The company itself was founded in 2010 by Morris Jette and Danny Auble, the original lead developers of Slurm. Danny Auble currently holds the position of CEO at SchedMD.
Nvidia stated that SchedMD will continue to operate Slurm as open-source and vendor-neutral software following the acquisition. The semiconductor company emphasized in its blog post that the technology qualifies as critical infrastructure for generative AI applications. Nvidia plans to maintain investments in Slurm and accelerate its accessibility across various systems. The two companies have collaborated for more than a decade prior to this deal. Terms of the acquisition were not disclosed, and Nvidia declined to provide additional comments beyond its blog post.
On the same day as the acquisition announcement, Nvidia introduced the Nemotron 3 family of open AI models. The company described this collection as the most efficient family of open models available for constructing accurate AI agents.
The Nemotron 3 lineup consists of three distinct variants tailored to different use cases:
- Nemotron 3 Nano: A small model optimized for targeted tasks.
- Nemotron 3 Super: A model engineered for multi-AI agent applications.
- Nemotron 3 Ultra: A model developed for more complicated tasks.
Jensen Huang, Nvidia’s founder and CEO, addressed the release in the company’s press release. He wrote, “Open innovation is the foundation of AI progress. 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.”
In the preceding week, Nvidia announced Alpamayo-R1, a new open-reasoning vision language model centered on research for autonomous driving. At that time, the company also incorporated additional workflows and guides for its Cosmos world models. These Cosmos models operate under an open-source permissive license and assist developers in utilizing them to create physical AI.
These developments form part of Nvidia’s efforts in recent months to strengthen its open-source and open AI portfolio. The company positions itself as a primary supplier of AI and software to robotics and self-driving vehicle companies developing the core technologies for their systems.
