Nvidia Unveils Open AI Models for Agents

Nvidia expands open-source presence with SchedMD acquisition and Nemotron 3 AI models.

As part of its expansion in the open-source domain, Nvidia acquired SchedMD and introduced the Nemotron 3 line of AI models.

SchedMD is the developer of the popular workload management system Slurm. The product was initially launched in 2002, and the company was founded eight years later by leading developers of the solution, Morris Jett and Danny Auble. The latter serves as CEO.

Nvidia has collaborated with SchedMD for over a decade and considers its technology a critical infrastructure for generative AI.

Additionally, Jensen Huang’s firm released the Nemotron 3 series of open neural networks, deemed the “most efficient” for creating accurate agents.

The series includes:

• Nemotron 3 Nano — a small LLM for targeted tasks;
• Nemotron 3 Super — designed for applications with multiple digital assistants;
• Nemotron 3 Ultra — intended for complex functions.

“Open innovation is the foundation of AI progress. With Nemotron, we are turning cutting-edge artificial intelligence into a platform that provides developers with the transparency and efficiency needed to create large-scale agent systems,” Huang stated in a press release.

Recently, Nvidia has been striving to strengthen its position in open-source.

In December, the company announced Alpamayo-R1 — an open visual thinking language model designed for autopiloting. The new tool is based on the “reasoning” Cosmos-Reason.

In October, Huang declared that artificial intelligence had reached a “success spiral.” According to him, significant improvements in neural networks lead to increased investment in the technology, further “boosting” the field.

In the third quarter, Nvidia’s revenue amounted to $57 billion, 62% more than the same period last year.