Nvidia and Palantir team up to run supply chains with AI, starting with Nvidia's own million-part operation

Details

Nvidia and Palantir are combining their platforms to automate supply chain operations, using Nvidia's own sprawling logistics as the initial deployment. A single Vera Rubin rack requires 1.3 million components and Nvidia's network spans thousands of suppliers globally — making it one of the more complex manufacturing chains in the industry and a credible stress test for the technology. Palantir is embedding Nvidia's open Nemotron models into its Foundry platform, allowing companies to fine-tune them on proprietary operational data without exposing that data to third-party providers. Nvidia cuOpt handles real-time scenario planning and optimization — identifying bottlenecks earlier, accelerating material allocation, and evaluating alternative sourcing paths when disruptions occur. Decisions and their outcomes feed back into the models, creating a continuous improvement loop. The full stack can run on a company's own hardware or in the cloud through the Sovereign AI OS reference architecture, with infrastructure partners Dell, Cisco, Rackspace, and Nebius. This gives enterprise buyers an on-premises option, consistent with Palantir CEO Alex Karp's stated position that companies should control their own AI rather than hand knowledge to third-party providers — a view Mistral founder Arthur Mensch has recently echoed. Palantir is presenting additional details at AIPCon 11. Manufacturing, pharma, and energy verticals are the stated initial target markets.