Palantir Technologies (NASDAQ: PLTR) announced a sovereign AI operating system reference architecture developed with NVIDIA that provides customers with a complete AI datacenter solution from hardware procurement to application deployment.

The Palantir AI OS Reference Architecture combines NVIDIA's Blackwell Ultra systems with eight GPUs and Spectrum-X Ethernet networking alongside Palantir's software suite, including AIP, Foundry, Apollo, Rubix, and AIP Hub. The architecture is based on NVIDIA Enterprise Reference Architectures and has been tested to run Palantir's complete software platform.

The system includes hardened Kubernetes infrastructure running Foundry services, unified management through Rubix and Apollo systems, and NVIDIA's AI Enterprise software stack with CUDA-X Libraries, Nemotron open models, and Magnum IO acceleration.

According to the company, the architecture addresses requirements for customers with existing GPU infrastructure, latency-sensitive workflows, data sovereignty needs, and high geographic distribution. The solution allows enterprises to maintain control over their data, AI models and applications while operating on NVIDIA infrastructure.

"From our first deployment with the United States government and in every deployment since, our software has had to meet the moment in the most complex and sensitive environments where customers must maintain control," said Akshay Krishnaswamy, Palantir's Chief Architect.

Justin Boitano, vice president of Enterprise AI Platforms at NVIDIA, stated that the architecture enables "industries and nations can turn data into intelligence with speed, efficiency, and trust."

The announcement was made based on information from a company press release.