NVIDIA and Palantir expand AI partnership into sovereign supply chains

NVIDIA and Palantir said on September 10, 2026, that they are extending their operational-AI partnership into critical supply chains. The first deployment is running inside NVIDIA’s own supply-chain organization.

For that deployment, NVIDIA’s supply-chain operations team worked with Palantir to build a Digital Supply Chain Intelligence command center in Palantir Foundry. It brings together the data inputs planners use for material-allocation decisions, according to a technical account published alongside the announcement.

The command center combines Palantir’s Foundry and Ontology systems with NVIDIA’s cuOpt engine and a customized version of NVIDIA’s Nemotron 3.5 Lightning model, the companies said. Ontology provides governed, unified context from NVIDIA’s supply-chain data. CuOpt, NVIDIA’s GPU-accelerated engine for large combinatorial optimization problems, runs a weekly material-allocation optimization and writes the proposed allocation back into the system.

Human planners remain in the decision loop. The companies’ technical description says their acceptances, edits and overrides — together with the resulting production outcomes — are written back into the Ontology. The model does not retrain itself in production.

The deployment builds on a relationship announced in 2025. That integrated operational-AI stack paired Palantir’s AIP platform and Ontology with NVIDIA accelerated computing, CUDA-X libraries and Nemotron models, and named Lowe’s as an early supply-chain user.

NVIDIA and Palantir said the sovereign-AI supply-chain stack can run on premises or in cloud environments, allowing organizations to align deployments with their data and operational requirements. That is the companies’ description of the product design, not the result of an independent test.

The companies have not disclosed a contract value, commercial terms, a general-availability date or external customers for the supply-chain stack. At publication, there was no independent verification of efficiency, resilience, reliability or cost-savings gains from NVIDIA’s internal deployment.