New Delhi: The global race to build powerful artificial intelligence infrastructure is pushing technology companies to rethink how AI systems are deployed. Governments, defence agencies and large enterprises now want AI platforms that run inside their own secure environments rather than relying only on public cloud services.
In that context, Palantir Technologies and NVIDIA have announced a new joint system designed to run AI workloads inside controlled environments. The companies revealed a Sovereign AI OS reference architecture, a blueprint that combines AI hardware, software and management tools to create what they describe as a turnkey AI data centre.
Palantir and NVIDIA launch sovereign AI operating system
The new system is built around Palantir AI OS Reference Architecture, often called AIOS-RA. The platform aims to provide a full stack AI infrastructure that organisations can deploy on their own hardware.
According to the announcement, the architecture combines NVIDIA accelerated AI computing with Palantir’s enterprise software stack.
The system includes several layers working together.
Key components include:
- NVIDIA Blackwell Ultra GPU systems for AI training and inference
- Spectrum X networking for high speed data transfer
- Palantir compute infrastructure running Foundry services
- Rubix zero trust Kubernetes management
- Apollo deployment and lifecycle automation tools
Together these elements form a production ready environment for running AI applications.
Built for governments and large enterprises
The companies say the architecture is designed for customers that require full control over their data and AI models.
That includes organizations dealing with sensitive information or strict regulatory environments.
Examples include:
- Government agencies
- Financial institutions
- Defense and intelligence organizations
- Large enterprises with data sovereignty requirements
The platform is designed to support deployments across on premise systems, edge infrastructure and sovereign cloud environments.
Akshay Krishnaswamy said the system builds on years of work with government customers. “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.”
He added, “Together with NVIDIA and building on many customers’ existing investments we are proud to deliver a fully integrated AI operating system that is optimized for NVIDIA accelerated compute infrastructure.”
NVIDIA highlights the role of AI infrastructure
The system runs on NVIDIA Blackwell Ultra GPUs, which are designed for training large AI models and running inference workloads.
The infrastructure stack also includes software tools such as CUDA libraries and NVIDIA AI Enterprise.
Justin Boitano said the growing demand for AI computing is changing how infrastructure is built, “AI is redefining the infrastructure stack demanding latency sensitive and data sovereign environments require a full stack architecture built from silicon to systems to software.”
He added, “By combining Palantir’s sovereign AI OS reference architecture with NVIDIA AI infrastructure industries and nations can turn data into intelligence with speed efficiency and trust.”
