- Start
- Sep 10, 202690% CONFIDENCEfrom the source
NVIDIA and Palantir Unveil a Sovereign AI Stack for Supply Chains at AIPCon 11
- On 10 September 2026 Palantir and NVIDIA announced a collaboration to bring sovereign AI to critical supply chains, starting with NVIDIA's own operations, and showcased it at Palantir's AIPCon 11[1][3]
- The stack brings NVIDIA Nemotron open models into Palantir Foundry and AIP, grounded in the Palantir Ontology, to give supply chain teams visibility, flag constraints and guide decisions while keeping control of proprietary data[1]
- NVIDIA works with its ecosystem to secure the 1.3 million parts in each Vera Rubin rack; the system starts with materials allocation decisions that affect how quickly parts move through the supply chain[1]
- NVIDIA's engineers write that cuOpt solves the weekly allocation problem to minimise Time of Ownership, while a post-trained Nemotron 3.5 Lightning learns from planners' decisions and rationales[2]
- The deployment runs on NVIDIA reference architectures and the jointly developed Palantir Sovereign AI Operating System Reference Architecture, supported by Dell Technologies and Cisco; it can run on premises with Cisco and Dell or in co-location and cloud with Rackspace and Nebius[1][4][5]
- Alex Karp called NVIDIA's supply chain arguably the most valuable, intricate and complex in the world; Jensen Huang said the partners are turning its operational graph into sovereign intelligence, from wafer to token[1]
- Palantir hosted AIPCon 11 on Thursday 10 September 2026, with NVIDIA, Cisco, Eaton, the FAA, L3Harris and others presenting[3]
Notable features
- Nemotron 3.5 Lightning, a mixture-of-experts model of 30 billion parameters with about 3 billion active, post-trained with NeMo tools inside a governed Palantir Autopilot lifecycle[2]
- 86.7% allocation-decision accuracy on NVIDIA's development benchmark, against 55.5% for the far larger Nemotron 3 Ultra and 17.5% for the base Lightning model[2]
- The LoRA post-training run finished on two NVIDIA B200 GPUs in minutes, so it can be repeated as planner feedback accumulates[2]
- NVIDIA cuOpt inside AIP for optimisation and scenario planning, reporting which constraints are binding each week[1][2]
References 584% CONFIDENCE
The first entry is always the pin's source. Overall confidence is a weighted average of how firmly each reference supports the start and end times used above; a reference counts half as much for every 180 days older than the newest.
- [1]90%nvidianews.nvidia.com/news/nvidia-and-palantir-bring-sovereign-intelligence-to-critical-supply-chainsnvidianews.nvidia.com· Posted Sep 23, 2026· Starts Sep 10, 2026 ✓· 21% of score
The release is dated September 10, 2026 ("Palantir Technologies Inc. and NVIDIA[2] today announced a collaboration") and says the stack "will be showcased in depth at Palantir's AIPCon 11 conference", which Palantir's Miami-datelined release says it hosted on Thursday, September 10, 2026; no time is given, so the pin is all-day.
- [2]92%From Wafer-Out to First Token: Codifying Supply Chain Expertise with Nemotron and Palantir Foundry (NVIDIA Technical Blog)developer.nvidia.com· Published Sep 10, 2026· 20% of score
NVIDIA's[1] same-day technical write-up of the deployment: the Digital Supply Chain Intelligence command center on Palantir Foundry, cuOpt solving the weekly material-allocation program, and a post-trained 30B Nemotron 3.5 Lightning that reached 86.7% allocation accuracy against 55.5% for Nemotron 3 Ultra.
- [3]80%Palantir Customers Set New Standard for AI at AIPCon 11 (Palantir release via Business Wire)stocktitan.net· Published Sep 10, 2026· 20% of score
Palantir's own release, datelined MIAMI (Business Wire), that it will host its eleventh AIPCon on Thursday, September 10, 2026 with NVIDIA[1][2] among the presenting customers showing how they optimise critical supply chains.
- [4]82%Sovereign AI for the Supply Chain, Built on Infrastructure You Control (Dell)dell.com· Published Sep 10, 2026· 20% of score
Dell's same-day post on the Palantir Sovereign AI OS Reference Architecture with NVIDIA[1][2] and Dell, describing the supply-chain AI stack announced that day and first built for NVIDIA's supply chain of over 1.3 million components per Vera Rubin rack.
- [5]78%Deploying AI You Control Doesn't Need to be So Hard (Cisco Blogs)blogs.cisco.com· Published Sep 10, 2026· 20% of score
Cisco's same-day announcement of its Palantir collaboration, with Cisco's Secure AI Factory with NVIDIA[1][2] as a preferred full-stack foundation for Palantir's Sovereign AI OS - the reference architecture the NVIDIA release names Cisco as supporting.
Suggest a correction
Something missing or wrong? Say it in your own words: a link that backs this pin up, a different start or end date and why, or a fact it lacks or gets wrong. The AI checks it against this pin's sources, searches for better ones, and adds any page that backs you up. The pin's own sources still count most.
