Nvidia and Palantir are expanding their partnership to build secure, sovereign AI systems for U.S. government agencies, combining Nvidia's industry-leading graphics processing units (GPUs) with Palantir's data management expertise. This week, Palantir announced a new open-source AI engine based on Nvidia's Nemotron models, designed to deliver artificial intelligence capabilities to federal agencies while keeping sensitive government data completely under their control.

What Makes This Partnership Different From Other Tech Collaborations?

The partnership addresses a critical challenge facing government agencies: how to harness the power of cutting-edge AI without sending sensitive data to cloud servers or external companies. The two companies are adapting Nvidia's Nemotron models for use in air-gapped environments, which are ultra-secure computer systems not connected to the internet. This approach allows government users to reap the productivity rewards of AI while maintaining complete control over classified and sensitive information.

Earlier this year, the companies announced the availability of a Palantir AI OS Reference Architecture, a turnkey AI datacenter that governments can deploy to develop their own internal AI capabilities. The system combines Nvidia's Blackwell Ultra platform and Blackwell GPUs with the company's CUDA library, a software toolkit that helps developers harness the raw computing power of GPUs for intensive AI applications. These systems are integrated with Palantir's Foundry, Apollo, and Artificial Intelligence Platform (AIP) software.

How Are These Companies Positioned to Dominate Government AI?

• Nvidia's GPU Dominance: Nvidia controls between 85% and 92% of the data center GPU market, where most AI processing occurs, giving the company an unmatched competitive advantage in providing the hardware backbone for government AI systems.
• Palantir's Data Intelligence Expertise: Palantir pioneered the concept of ontology, overlaying an AI-centric dashboard across company information systems to enable real-time, data-driven decision-making, which sits at the heart of its wildly successful Artificial Intelligence Platform.
• Complementary Strengths: Palantir has deep relationships with enterprise and government customers who need AI solutions, while Nvidia supplies the bleeding-edge processors required to actually run those solutions at scale.

The partnership is particularly significant because it mirrors how open-source software shaped the early internet through collaboration between the Defense Advanced Research Projects Agency (DARPA) and top-ranked universities. By making these models open-source and adaptable, the companies are positioning themselves as essential infrastructure providers for a new era of government AI.

What Do the Financial Numbers Tell Us About Their Growth?

Both companies are experiencing explosive growth that validates their market position. In the first quarter of 2026, Palantir generated revenue of $1.63 billion, up 85% year over year, marking the company's highest-ever year-over-year growth rate and the 11th consecutive quarter of accelerating revenue growth. The company's profitability was equally impressive, with adjusted earnings per share surging 154%.

Perhaps more telling, Palantir's remaining performance obligation, commonly called backlog, surged 134% year over year to $4.45 billion. This metric matters because it represents future revenue already committed by customers. Additionally, Palantir's net dollar retention rose to 150%, meaning current customers are spending 50% more than they were a year earlier.

Nvidia's numbers are equally staggering. For its fiscal 2027 first quarter ending April 26, Nvidia generated record revenue that surged 85% year over year to $81.6 billion, which drove adjusted earnings per share that soared 140% to $1.87, marking the 14th consecutive quarter of sequential revenue growth. Nvidia CEO Jensen Huang has stated that the company has "visibility" to $1 trillion in revenue from the company's Blackwell and Vera Rubin chips over the next couple of years.

Steps to Understanding the Government AI Opportunity

• Recognize the Security Imperative: Government agencies cannot store classified information on public cloud servers, creating a unique market for on-premises AI systems that Nvidia and Palantir are specifically designed to serve.
• Understand CUDA's Competitive Moat: CUDA is a library of software tools that helps developers harness GPU computing power, and it gives Nvidia a vast competitive advantage because competitors cannot easily replicate this ecosystem of tools and optimizations.
• Track the Backlog Metric: Palantir's $4.45 billion backlog is a leading indicator of future revenue, suggesting that government demand for sovereign AI infrastructure is already locked in and will drive growth for years to come.

The partnership between Nvidia and Palantir represents more than just a business collaboration; it reflects a broader shift in how governments approach artificial intelligence. Rather than relying on commercial cloud providers, agencies are building internal AI capabilities that keep sensitive data secure while still accessing the latest AI technology. This trend is likely to accelerate as more government agencies recognize the competitive advantage of having AI systems they fully control.

Both companies are trading at valuations that analysts consider attractive for their growth rates. Palantir is selling for 131 times earnings, compared to Nvidia's more reasonable multiple of 30. However, when using the price-to-earnings-to-growth ratio, a more appropriate metric for high-growth stocks, both companies show multiples below 1.0, which is the standard threshold for undervalued stocks.