At the India AI Summit, Larsen & Toubro (L&T) unveiled a proposed venture under the Government of India’s IndiaAI Mission to build sovereign, scalable gigawatt-scale AI factory infrastructure powered by NVIDIA technologies.
The initiative is designed to strengthen India’s position as a global AI powerhouse by creating production-grade AI capacity anchored within the country’s digital and industrial ecosystem.
The proposed venture will combine L&T’s expertise in engineering, infrastructure development and execution with NVIDIA’s advanced AI infrastructure stack. This includes NVIDIA GPUs and CPUs, high-performance networking, NVIDIA-accelerated storage platforms from leading providers, the NVIDIA AI Enterprise software stack, and validated reference architectures. The integrated model aims to accelerate secure and large-scale AI adoption across industries.
Under the framework of the IndiaAI Mission, the venture seeks to establish sovereign AI infrastructure that enables critical data, AI models and workloads to be built, trained and deployed entirely within India. While anchored domestically, the platform will remain interoperable with global ecosystems, positioning India as a strategic AI hub for domestic enterprises, global hyperscalers, cloud providers and international off-takers.
A key component of the plan is the development of a gigawatt-scale AI data centre factory capable of supporting high-density, next-generation AI workloads. The infrastructure is intended to provide AI-ready capacity that allows enterprises to scale operations efficiently and sustainably.
As part of the expansion roadmap, L&T plans to scale NVIDIA GPU cluster deployment at its Chennai data centre campus up to 30 MW capacity within its 300-acre gigawatt-scalable campus. Additionally, a new 40 MW data centre in Mumbai, currently under execution, will further strengthen the company’s AI infrastructure footprint.
The AI factory model is structured to deliver advanced AI services to global off-takers, hyperscalers and India Inc across sectors such as manufacturing, infrastructure, energy, financial services, healthcare and public services. The goal is to help organisations transition from pilot-stage experimentation to full production-scale AI deployment.
The infrastructure will offer standardised, enterprise-grade AI capabilities designed for predictable performance, enhanced security and faster time-to-value for industrial and services use cases. By enabling sovereign cloud-based AI deployment, the venture also supports internal transformation across L&T and its group companies.
This includes initiatives such as LTTS’s Lights-Out Factory framework leveraging NVIDIA Omniverse libraries, LTM’s Blueverse platform, LTFS’s agentic AI deployment and L&T’s internally developed AI agents. Together, these efforts are aimed at building a self-sustaining innovation ecosystem powered by sovereign AI infrastructure in India.