Pranav Hotkar
New Delhi, India - February 19, 2026 - Larsen & Toubro has partnered with NVIDIA to develop sovereign, gigawatt-scale AI data center infrastructure under the government’s IndiaAI Mission, marking one of the country’s largest largestplanned computing capacity expansions.
The collaboration will build what the companies describe as an “AI factory” platform designed to train and deploy large artificial-intelligence models within India while remaining interoperable with global systems. The project targets enterprise, public-sector and research workloads across manufacturing, financial services, healthcare, and energy.
Initial deployments include expansion of L&T’s existing AI compute facility in Chennai to roughly 30 MW and construction of a new 40 MW site in Mumbai. The long-term roadmap scales beyond individual campuses toward gigawatt-level capacity, positioning the infrastructure as a national compute backbone.
According to company statements, the facilities will integrate NVIDIA accelerated computing technology, including GPUs, CPUs, networking, and AI software, into a unified architecture optimized for high-density workloads. The design emphasizes localized data processing, high-bandwidth interconnects and energy-efficient operation required for large-model training clusters.
L&T Chairman and Managing Director S. N. Subrahmanyan said the partnership aims to translate AI capability into measurable economic impact, adding that domestic compute availability would help industries deploy advanced analytics and automation at scale. NVIDIA Chief Executive Jensen Huang stated the collaboration establishes foundational infrastructure to power India’s next phase of industrial and digital growth.
Industry analysts view the project as part of a broader global shift toward sovereign AI infrastructure, where nations seek local capacity for training and deploying critical models instead of relying entirely on overseas hyperscale platforms. High-density AI facilities require significantly more power and cooling than traditional cloud data centers, driving new campus-scale developments rather than incremental expansion of existing sites.
The initiative aligns with India’s policy push to strengthen domestic computing capacity, support local AI model development, and reduce dependence on cross-border processing for sensitive workloads.
If executed as planned, the partnership would place India among a small group of regions building national-scale AI infrastructure, signaling a transition from cloud consumption to locally anchored AI production capability.
