NVIDIA and Palantir Technologies are integrating NVIDIA Nemotron open models with Palantir Foundry, Artificial Intelligence Platform (AIP) and Ontology to run AI-assisted supply chain planning across NVIDIA’s operations. The system models supply constraints and materials-allocation trade-offs using proprietary data while keeping control of the models, data and deployment environment within the organization.
NVIDIA’s supply chain spans millions of parts, thousands of suppliers and a global manufacturing network. Producing rack-scale AI systems requires availability of compute, memory, networking, power, cooling and mechanical components. Each NVIDIA Vera Rubin rack contains about 1.3 million parts.
Palantir Ontology maps supply chain relationships and operating context for the deployment. Nemotron models use that context with data in Foundry and AIP to support decisions across component supply, manufacturing and production planning.
“NVIDIA has arguably the most valuable, intricate and complex supply chain in the world. Our sovereign stack, powered by Nemotron models and Ontology, is delivering capabilities that exceed the frontier while providing alpha protection qualities unavailable otherwise,” said Alex Karp, cofounder and CEO of Palantir Technologies. “We are very proud of our partnership and its cornerstone role in the sovereign AI revolution.”
“Supply chains are the operating system of the physical economy, and AI factories are among the most complex systems ever built,” said Jensen Huang, founder and CEO of NVIDIA. “From wafers and components to manufacturing, systems and customer delivery, hundreds of companies and trillions of dollars of global economic activity come together to deliver AI infrastructure. NVIDIA and Palantir are transforming this vast operational graph into sovereign intelligence – combining NVIDIA Nemotron models with Palantir’s Ontology to reason, plan and orchestrate the journey from wafer to token.”
Post-Training Nemotron for Supply Chain Decisions
Palantir customers can post-train Nemotron models through Foundry and AIP. NVIDIA NeMo Data Libraries prepare and augment the data before model customization. Organizations can adapt the models to their supplier networks, operating constraints and decision criteria.
Within Palantir AIP, NVIDIA cuOpt handles optimization and scenario planning. Teams can model supply constraints, compare allocation options and examine how decisions affect downstream production. Post-trained Nemotron models can recommend actions, explain trade-offs and flag emerging risks, while supply chain specialists retain final decision authority.
NVIDIA is initially using the system for materials allocation. Supply chain teams can identify potential constraints, compare alternatives and allocate parts based on their effect across production. The deployment also gives teams a view of supply conditions across NVIDIA’s network.
Palantir Autopilot connects production outcomes and user feedback with NVIDIA NeMo AutoModel and NeMo RL libraries. The feedback loop compares recommendations with operating results and feeds those results into model training. This process captures operational knowledge and refines models used in subsequent supply chain decisions.
Deployment Architecture and Infrastructure Options
The deployment runs on NVIDIA reference architectures and the Palantir Sovereign AI Operating System Reference Architecture (SAIOS). Dell and Cisco support the reference architecture. Organizations can deploy the stack on premises using systems from companies including Cisco and Dell, or through colocation and cloud environments from Rackspace and Nebius.
NVIDIA and Palantir plan to extend the architecture to supply chain operations in manufacturing, energy, health care, automotive, aerospace, agriculture, pharmaceuticals, retail and government. The companies will present the NVIDIA deployment and related industry applications at Palantir’s AIPCon 11 conference.
Source: NVIDIA
About NVIDIA
NVIDIA, founded in 1993 and headquartered in Santa Clara, CA, designs and manufactures graphics processing units, systems on chips, networking hardware, and AI intelligence software such as CUDA. Its products serve industries including gaming, data centers, autonomous vehicles, professional visualization, robotics, health care, and energy. The company introduced the GPU in 1999 and later expanded into accelerated computing and AI infrastructure. In gaming, its GPUs support high-performance rendering, while in AI and high-performance computing, its systems provide the infrastructure for training and deploying large-scale models. NVIDIA also develops tools for robotics and autonomous driving.
About Palantir Technologies
Palantir Technologies is a software company founded in 2003 and headquartered in Aventura, FL. The company develops software for integrating, managing and analyzing data across complex operations. Its main products include Foundry, Gotham, Apollo and the Artificial Intelligence Platform (AIP). Foundry supports data management, analytics and operational workflows for commercial organizations. Gotham serves government and defense users working with large, distributed data sets. Apollo manages software deployment across cloud, on-premises and disconnected environments. AIP connects AI models with organizational data and operating workflows. Palantir serves government agencies and companies in manufacturing, health care, energy and transportation. Engineers, data scientists, operators and information technology teams use its software. Palantir reported about 1,000 customers worldwide. The company also reported about 4,000 full-time employees globally.
