L&T Lands Mega Order to Build India's Largest NVIDIA B300 AI Factory for Together AI

Larsen & Toubro (L&T) announced on 13 August 2026 that it has secured a "mega" order to build what it describes as India's largest single-cluster artificial-intelligence computing facility — an NVIDIA B300 "AI Factory" housing 10,000 GPUs — for the US-based cloud company Together AI. In L&T's own contract-classification language, a "mega" order is one valued between ₹10,000 crore and ₹15,000 crore, placing this among the largest AI-infrastructure deals yet awarded to an Indian engineering company and marking L&T's formal entry into the AI-factory business.

What is being built

The facility will be delivered by Vyoma.AI, an L&T company, through its AI-infrastructure subsidiary LTN Compute. At its heart sits a single cluster of 10,000 NVIDIA B300 GPUs — the latest generation of NVIDIA's data-centre accelerators — wired together with high-throughput networking, ultra-low-latency interconnects and parallel storage into an integrated "AI Factory."

The cluster will be hosted at Vyoma's data-centre campus in Chennai, which L&T describes as a gigawatt-scale AI infrastructure site. Phase 1 of the campus is designed for 250 MW of IT load, with power-infrastructure readiness of 150 MVA already in place. Once live, the platform is intended to run large-scale inference, fine-tuning and model-training workloads as an "AI-native cloud."

Why an Indian builder for a US cloud

The customer is Together AI, a US-headquartered AI cloud platform provider that rents out high-performance GPU capacity to companies building and serving large language models and other generative-AI systems. Rather than concentrate all of its compute in existing US or European hubs, Together AI is turning to India for a dedicated, single-tenant cluster — a signal that the country is increasingly viewed as a competitive location for frontier AI compute, thanks to falling power costs, available land and a deepening engineering talent pool.

For L&T, the deal fits a broader pivot. The 80-year-old engineering-and-construction conglomerate has been positioning its digital and data-centre arms to capture demand from India's AI build-out, and the Together AI order gives it an anchor customer and a reference deployment at the very top end of the market.

The sovereignty and scale angle

L&T has framed the project around two ideas: scale and sovereign infrastructure. The venture aims to keep data, models and workloads resident within India while remaining compatible with global systems — an increasingly important consideration for enterprises and governments wary of routing sensitive AI workloads through overseas data centres.

"Our deployment of an NVIDIA B300 AI Factory for Together AI marks a significant milestone in L&T's Gigawatt AI Infrastructure Mission and reinforces our commitment to making India a global hub for next-generation AI infrastructure," said S.N. Subrahmanyan, Chairman and Managing Director of L&T.

The "Gigawatt AI Infrastructure Mission" branding points to L&T's ambition to build out multiple hundreds of megawatts — ultimately gigawatts — of AI-ready capacity over the coming years, of which the Chennai campus is the first flagship.

Context: India's compute build-out

The order lands in the middle of an unprecedented surge in AI-infrastructure commitments across India. Global hyperscalers and domestic conglomerates alike have pledged tens of billions of dollars toward data centres and accelerated computing, while the government's IndiaAI Mission has been assembling a shared GPU pool and backing sovereign foundation-model efforts. NVIDIA has separately tied up with several Indian groups to seed domestic AI supercomputing.

Against that backdrop, a 10,000-GPU B300 cluster is significant not just for its size but for what it represents: an Indian firm taking on the full engineering, power and operations stack for a hyperscale AI deployment, rather than simply leasing space or importing turnkey systems.

What to watch

Key questions now turn to execution and timelines. Building and commissioning a gigawatt-scale campus — securing power, cooling, networking and the GPUs themselves amid tight global supply — is a formidable undertaking, and L&T has not published a completion date for the full cluster. Delivery cadence, the ramp from Phase 1's 250 MW toward the campus's gigawatt ceiling, and whether L&T can convert this anchor deal into repeat AI-factory orders will determine how much of India's AI-infrastructure wave the company ultimately captures.

Sources

• Larsen & Toubro press release, "Larsen & Toubro Secures Mega Order… to build India's largest NVIDIA B300 AI Factory," 13 August 2026: https://www.larsentoubro.com/pressreleases/2026/2026-08-13-larsen-toubro-secures-mega-order-as-part-of-a-strategic-partnership-with-together-ai-to-build-india-s-largest-nvidia-b300-ai-factory
• Business Standard, "L&T to set up India's largest Nvidia B300 AI factory for Together AI," 13 August 2026: https://www.business-standard.com/companies/news/l-and-t-to-set-up-nvidia-b300-ai-factory-for-together-ai-126081301000_1.html
• The Free Press Journal, "L&T Wins ₹10,000-15,000 Crore Mega Order, To Build NVIDIA's 10,000-GPU AI Factory In Chennai," August 2026: https://www.freepressjournal.in/business/lt-wins-10000-15000-crore-mega-order-to-build-nvidias-10000-gpu-ai-factory-in-chennai
• w.media, "L&T partners with Together AI to deploy NVIDIA B300 infrastructure at Chennai AI factory," August 2026: https://w.media/lt-partners-with-together-ai-to-deploy-nvidia-b300-infrastructure-at-chennai-ai-factory/