"AI agents must be able to cooperate, coordinate, and execute across large contexts and long time periods, and this, says Nvidia, demands a new type of infrastructure, one that is open. The company says it has the answer with its new Nemotron 3 family of open models. Developers and engineers can use the new models to create domain-specific AI agents or applications without having to build a foundation model from scratch."
"Nemotron 3 features what Nvidia calls a "breakthrough hybrid latent mixture-of-experts (MoE) architecture". The model comes in three sizes: Nano: The smallest and most "compute-cost-efficient," intended for targeted, highly-efficient tasks like quick information retrieval, software debugging, content summarization, and AI assistant workflows. The 30-billion-parameter model activates 3 billion parameters at a time for speed and has a 1-million-token context window, allowing it to remember and connect information over multi-step tasks."
Nvidia positions Nemotron 3 as an open infrastructure to enable AI agents that cooperate, coordinate, and execute across large contexts and long timeframes. The Nemotron 3 family uses a hybrid latent mixture-of-experts (MoE) architecture and ships in Nano, Super, and Ultra sizes optimized for different compute and reasoning needs. Nvidia will release most training data and reinforcement learning libraries to support building domain-specific agents without training foundation models from scratch. Nano targets efficient retrieval and assistant workflows with a 1-million-token context window. Super and Ultra scale to hundreds of billions of parameters to support multi-agent collaboration, deep research, and complex reasoning with low latency.
Read at InfoWorld
Unable to calculate read time
Collection
[
|
...
]