NVIDIA and Microsoft Advance Agentic AI from Cloud to PC

NVIDIA and Microsoft join forces to accelerate Artificial Intelligence-driven scientific discovery, launching a new extensible platform to revolutionize research workflows.

NVIDIA and Microsoft have announced a significant deepening of their partnership aimed at advancing agentic Artificial Intelligence-powered applications across both cloud and personal computing environments. Unveiled during Microsoft Build, the new Microsoft Discovery platform is designed to transform how researchers approach the entire discovery process, leveraging agentic Artificial Intelligence to streamline workflows and foster quicker breakthroughs. The initiative targets research and development functions across numerous industries, promising to accelerate the timeline to market for new products and expand the capacity of scientists to make novel discoveries more efficiently.

Microsoft Discovery will feature integrations with key NVIDIA technologies, notably the NVIDIA ALCHEMI NIM microservice, which is optimized for Artificial Intelligence inference in chemical simulations. This integration is expected to drive faster insights in materials science by enabling rapid property prediction and candidate recommendation. In addition, the platform will incorporate NVIDIA BioNeMo NIM microservices to expedite Artificial Intelligence model development in the field of drug discovery, significantly enhancing the speed and scale at which research teams can work.

Testing has already demonstrated the platform’s potential: Microsoft researchers leveraged Microsoft Discovery to identify a promising novel coolant prototype suitable for immersion cooling in data center environments, completing the process in under 200 hours. This timeline represents a substantial improvement compared to traditional research methods, which could take months or years. The collaboration signals a move toward an era where cross-company partnerships and advanced Artificial Intelligence technologies underpin accelerated scientific progress from the cloud down to individual PCs.

77

Impact Score

Introducing Mistral 3: open artificial intelligence models

Mistral 3 is a family of open, multimodal and multilingual Artificial Intelligence models that includes three Ministral edge models and a sparse Mistral Large 3 trained with 41B active and 675B total parameters, released under the Apache 2.0 license.

NVIDIA and Mistral Artificial Intelligence partner to accelerate new family of open models

NVIDIA and Mistral Artificial Intelligence announced a partnership to optimize the Mistral 3 family of open-source multilingual, multimodal models across NVIDIA supercomputing and edge platforms. The collaboration highlights Mistral Large 3, a mixture-of-experts model designed to improve efficiency and accuracy for enterprise artificial intelligence deployments starting Tuesday, Dec. 2.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.