AI

HCLTech & Sarvam AI to Build $1.48B AI Data Center in Odisha

By Vikramaditya Sen | Published July 25, 2026 | 6 min read

HCLTech & Sarvam AI to Build $1.48B AI Data Center in Odisha

India doubles down on sovereign AI as one of the country's biggest AI infrastructure projects takes shape.

IT major HCLTech and generative AI pioneer Sarvam AI have officially announced a joint $1.48 Billion (₹12,300 Crore) venture to construct one of India's largest mega AI data center campuses in Odisha. The strategic initiative marks a pivotal milestone in India's push for Sovereign Compute Infrastructure, delivering high-density GPU acceleration tailored specifically for multilingual Indian Large Language Models (LLMs) and high-performance enterprise workloads.

Located near Bhubaneswar's emerging InfoValley tech corridor, the greenfield data center facility is designed for an initial capacity of 120 MW, with plans to scale to 300 MW over the next three years.

Key Project Specifications & Investment Metrics

Metric / DimensionProject DetailStrategic Significance
Total Investment$1.48 Billion (₹12,300 Cr)One of Asia-Pacific's largest sovereign AI infrastructure investments
Initial Compute Power120 MW High-Density Liquid CooledScalable to 300 MW dedicated AI workload capacity
Primary Industry PartnersHCLTech & Sarvam AIEnterprise IT integration paired with domestic AI model expertise
Target GPU Deployment32,000+ Next-Gen AcceleratorsCustom clusters optimized for LLM pre-training & fine-tuning
PUE (Power Usage Effectiveness)Target < 1.25Powered by 70% renewable hydro & solar power from Odisha grid
Expected Operational DateQ2 2027 (Phase 1 launch)Accelerated 18-month construction & commissioning timeline

Accelerating Sovereign AI for 1.4 Billion Citizens

The partnership fuses HCLTech’s global infrastructure management expertise with Sarvam AI’s cutting-edge Indian language models (such as Sarvam-1 and Shunya). By housing training compute directly on Indian soil, the Odisha data center ensures absolute data sovereignty, low latency for domestic enterprises, and strict compliance with national privacy regulations.

"To build true AI independence, India requires world-class compute infrastructure designed for native languages and localized enterprise security,"
stated Pratyush Kumar, Co-founder of Sarvam AI. "This $1.48 billion Odisha facility gives Indian researchers, startups, and enterprises the compute power of Silicon Valley within home borders."

Strategic Advantages of Odisha as an AI Infrastructure Hub

Industry experts highlight several structural advantages that led to Odisha being chosen for this mega facility:

1. Abundant Clean Energy: Odisha’s vast hydroelectric reservoirs and expanding solar infrastructure supply uninterrupted green power, crucial for meeting net-zero carbon goals.
2. Subsea Cable Connectivity: Direct coastal access to landing stations enables ultra-low latency routing to South East Asia and the Middle East.
3. State Policy Support: The Odisha Electronics & IT Policy 2026 offers custom power subsidies, single-window environmental clearances, and stamp duty waivers for mega data centers.

Impact on India's Broader Tech Ecosystem

This mega data center investment builds upon earlier momentum in India's Spacetech & Deeptech Boom and aligns with Enterprise AI's shift to Autonomous Automation. As global tech giants scramble for AI compute, India's sovereign data centers guarantee domestic startups access to affordable, high-end GPU capacity.

Related Intelligence Reports

Sarvam AI Partners with IBM to Advance Sovereign AI Solutions for India's Government and Regulated Sectors
AI

Sarvam AI Partners with IBM to Advance Sovereign AI Solutions for India's Government and Regulated Sectors

By Dr. Ananya Roy · Aug 1, 2026

HCLTech Bets Big on AI Infrastructure with ₹3,500 Crore Investment
Tech

HCLTech Bets Big on AI Infrastructure with ₹3,500 Crore Investment

By Vikram Mehta · Jul 14, 2026

Indian AI Startups Target the “Model Routing” Opportunity: Slashing Enterprise Inference Costs and Latency
AI

Indian AI Startups Target the “Model Routing” Opportunity: Slashing Enterprise Inference Costs and Latency

By Elena Rostova · Sep 22, 2026