Nvidia and Palantir deploy sovereign AI for critical supply chains

Nvidia and Palantir Technologies are rolling out a sovereign AI platform for complex supply chains, starting with Nvidia's own operations. The system combines Nvidia Nemotron open models with Palantir’s Foundry and AIP to improve visibility and decision speed while keeping proprietary data under the company’s control. The move signals how enterprise AI is increasingly moving into factory-floor and logistics decision-making, a trend directly relevant to European industrial and manufacturing firms.

Bottom line — Each Nvidia Vera Rubin rack requires 1.3 million parts; the AI stack aims to shorten the wafer-to-token journey.

Go deeper (7)

• The AI stack uses Nemotron open models post-trained on Nvidia’s proprietary operational data within Palantir Foundry, according to the companies’ press release.
• Nvidia reported that human employees still provided key insights — such as emails with partners, severe weather forecasts, and geopolitical events — that the AI struggled to incorporate, per the Jerusalem Post.
• The system is designed to create a feedback loop via NeMo AutoModel and Palantir Autopilot, where operational results improve specialized models over time, the press release states.
• The deployment can run on-premises, in colocation facilities, or in the cloud, supported by infrastructure partners including Dell Technologies and Cisco, per the companies.
• Organizations in agriculture, manufacturing, pharmaceuticals, retail, technology and government can use the AI stack to optimise their own supply chains, the companies said.
• Palantir CEO Alex Karp called the partnership a 'cornerstone role in the sovereign AI revolution,' as quoted in the announcement.
• The European tech press notes the project highlights demand for AI systems that keep proprietary industrial data under an organisation’s control, relevant for European manufacturers.

Read the reporting