NVIDIA Expands Open Models for European Industries

NVIDIA is launching a comprehensive range of open models designed for language processing, robotics, biology, and quantum computing, enabling developers, researchers, and companies to build applications tailored to local languages and industries. The newcomers in this portfolio include the NVIDIA Nemotron 3.5 models, Nemotron Retriever models for document analysis, and the Kumo family of models for structured business data, alongside tools for robotics and quantum applications.

The NVIDIA Nemotron models provide a platform for developing text and image processing agents that can effectively perform complex tasks. This allows organizations to build applications based on their proprietary knowledge, with support from industry partners such as Black Forest Labs and Mistral. Well-known companies like Salesforce and Accenture are participating in integrating these models into their systems and workflows, which helps them create customized solutions.

Moreover, NVIDIA Kumo enables the analysis of structured business data to generate predictive insights faster. This saves time and reduces the complexity of setting up different machine learning pipelines for various tasks. Companies like Coinbase and Databricks are already benefiting from the Kumo models to accelerate AI applications.

Furthermore, NVIDIA Cosmos provides an environment for developing robots and autonomous vehicles. With tools like the Cosmos Data Factory, developers can create high-quality training datasets to enhance the accuracy of their models. New models such as NVIDIA Alpamayo 2 Edge are focused on real-time scenario analyses for autonomous driving. Finally, NVIDIA's open Ising models accelerate the development of quantum technology and are more widely available through platforms like AWS. These steps mark NVIDIA's commitment to supporting collaboration and innovation across various industries in Europe.

Read the full article from Nvidia.