Nvidia has announced a number of partnerships with leading infrastructure and technology providers in India to help build sovereign artificial intelligence (AI) capabilities in the Indian subcontinent.
At the AI Impact Summit in New Delhi this week, the chip giant unveiled plans to deploy tens of thousands of its latest Blackwell graphics processing units (GPUs) across the country in support of the government-backed IndiaAI mission to accelerate AI innovation by improving data quality, developing AI talent, and improving access to AI infrastructure.
For example, Nvidia is working with engineering giant Larsen & Toubro (L&T) to build a gigawatt-scale AI data center network. These facilities are intended to store critical data and model training within India’s borders, which is a key requirement for governments and regulated industries.
“AI is driving the creation of the greatest infrastructure in human history. Everyone will use AI, every company will leverage AI, and every country will build AI,” said Nvidia CEO Jensen Huang. “Together with L&T, an 88-year-old engineering and nation-building leader, we are laying the foundation for a world-class AI infrastructure that will fuel India’s growth and help realize India’s complete vision of AI.”
Meanwhile, cloud service provider Yotta has deployed over 20,000 Nvidia Blackwell Ultra GPUs to power its Shakti Cloud hosted in data centers in Navi Mumbai and Greater Noida. E2E Networks also plans to launch a Blackwell cluster on an AI and machine learning platform hosted at L&T Vyoma data center in Chennai.
The AI infrastructure in these data centers will serve the needs of model builders, startups, researchers, and enterprises building and deploying AI applications in India, as well as supporting efforts such as BharatGen, a government-backed project to create a multilingual and multimodal AI ecosystem.
As part of the BharatGen project, a new 17 billion parameter expert mixture model was built from scratch using Nvidia’s NeMo framework for pre-training and the NeMo RL library for post-training. The open source model is designed to power applications across public services, agriculture, security, and cultural preservation.
Additionally, Nvidia is helping India’s $250 billion IT services sector move beyond generative AI chatbots to agent-based AI systems that can reason and execute complex workflows.
Leading global system integrators such as Infosys, TCS, Wipro, and Tech Mahindra announced that they will adopt Nvidia’s AI Enterprise software to build AI agents for their customers around the world.
Wipro introduced a new system developed for US health insurance companies. The system can now handle 42% of incoming calls with less than 200 milliseconds of latency. Meanwhile, Infosys has integrated Nvidia’s stack to create a 2.5 billion parameter model for agent development, code generation, refactoring, and software engineering workflows.
“Agent AI is reshaping India’s technology industry and delivering service breakthroughs around the world,” Nvidia said in a blog post, noting that “technology leaders are leveraging Nvidia Nemotron’s open models, data, and software to build and deploy agentic and generative AI to accelerate productivity and efficiency.”
The summit also highlighted India’s growing role as a manufacturing hub, with industry giants Reliance Industries and Tata Motors adopting Nvidia’s Omniverse platform to build physical AI applications.
For example, Reliance is combining Siemens’ digital twin technology with the Omniverse Library to enable faster and more accurate simulation and plant design for next-generation gigafactories. Meanwhile, Tata Motors is using digital twins and other AI capabilities built on the Omniverse library to transform camera feeds into intelligent sensors for automated quality checks and real-time safety compliance.
Nvidia said these partnerships are aimed at “bringing physical AI to factories, warehouses, and infrastructure,” which will help Indian manufacturers reduce lead times and improve operational efficiency through high-fidelity simulation.
