India strengthens sovereign AI infrastructure with one of Asia’s largest NVIDIA Blackwell HGX B300 superclusters, set to go live by August 2026.
Yotta Data Services has announced plans to deploy 20,736 liquid-cooled NVIDIA Blackwell Ultra GPUs, forming one of Asia’s largest AI superclusters and positioning India as a major global AI infrastructure hub. The deployment represents a capital investment exceeding $2 billion and is expected to go live by August 2026.
In parallel, NVIDIA will establish one of the Asia-Pacific region’s largest NVIDIA DGX Cloud clusters within Yotta’s Blackwell HGX B300 supercluster under a four-year engagement valued at over $1 billion. The agreement significantly expands an existing collaboration, as NVIDIA DGX Cloud has already been utilizing Yotta’s GPU infrastructure over the past year.
One of Asia’s Largest AI Superclusters Built on Blackwell Ultra
The AI supercluster will be deployed at Yotta’s 60 MW D2 hyperscale data center at its Greater Noida campus, which is scalable to 250 MW. Additional capacity will be supported by Yotta’s Navi Mumbai campus, scalable up to 2 GW. The infrastructure is engineered to support trillion-parameter foundation model training and high-throughput inference workloads capable of processing millions of simultaneous prompts.
Built on NVIDIA reference architecture, the supercluster integrates 800 Gbps NVIDIA Quantum-X800 InfiniBand networking, advanced liquid-cooling systems, and over 40 petabytes of high-performance parallel file-system storage. The platform is designed to deliver high energy efficiency, reduced cost per token, and accelerated time-to-market for AI model development.
With this deployment, India joins a select group of geographies capable of hosting frontier-scale AI infrastructure, marking a pivotal step in strengthening sovereign AI compute capacity and reducing reliance on offshore hyperscalers.
APAC’s Largest DGX Cloud Cluster in India
Under the multi-year agreement, NVIDIA will establish one of APAC’s largest DGX Cloud clusters within Yotta’s Blackwell Ultra supercluster. The expansion reflects rising regional and global demand for AI training and inference capacity and signals growing India–U.S. technology collaboration.
The move also aligns with broader global shifts in AI compute supply chains, where high-performance AI infrastructure is increasingly distributed across trusted regions. India’s emergence as a strategic AI compute node reinforces shared priorities around secure, resilient, and high-performance AI ecosystems.
Powering IndiaAI Mission and Sovereign AI Development
Yotta will allocate over 10,000 NVIDIA B300 GPUs from the supercluster to support the IndiaAI Mission, enabling sovereign foundation model development, academic research, AI startups, and population-scale public AI platforms. This ensures that domestic AI innovation progresses alongside international AI workloads hosted within India.
Through its Shakti Studio AI platform, Yotta is integrating NVIDIA Nemotron open models, NVIDIA NIM microservices, and the full NVIDIA AI Enterprise software suite. Indian developers will gain access to open model weights, training datasets, and model recipes, enabling transparent fine-tuning and customization for India-specific use cases.
By combining Blackwell Ultra GPU infrastructure with open AI architectures and enterprise-grade software, Yotta aims to enable startups, enterprises, and public institutions to build secure, production-ready generative AI applications at scale.
Strategic Capital Commitment and Long-Term Scale
The combined capital commitments—over $2 billion for Blackwell Ultra infrastructure deployment and more than $1 billion in contracted DGX Cloud capacity—provide long-term demand visibility and reflect sustained appetite for high-performance AI compute in India and the broader Asia-Pacific region.
Darshan Hiranandani, Co-Founder and Chairman of Yotta Data Services, emphasized that AI infrastructure is becoming foundational economic infrastructure, reinforcing India’s role in the global AI value chain. Sunil Gupta, Co-Founder, MD and CEO of Yotta, noted that combining Blackwell Ultra GPUs with open models such as Nemotron will enable India to build sovereign, globally competitive AI systems.
NVIDIA co-founder and CEO Jensen Huang stated that India is emerging as one of the world’s most important AI markets, driven by deep technical talent and an ambitious national vision. He highlighted that Yotta’s Blackwell Ultra supercluster would enable frontier-scale AI model training and population-scale AI deployment within India.
Scaling Toward 80,000+ GPUs and Beyond
Yotta currently operates over 10,000 NVIDIA GPUs in production, with an additional 8,000 GPUs expected to go live within the next quarter. The deployment of 20,736 Blackwell Ultra GPUs by August 2026 marks a significant expansion phase. The company has outlined a roadmap to scale beyond 80,000 NVIDIA GPUs by FY27, with long-term plans to support over one million GPUs as India’s AI ecosystem matures.
With hyperscale data center infrastructure, green energy sourcing, extra-high-voltage substations, and vertically integrated cloud and managed services capabilities, Yotta is building a structurally significant AI compute hub within the global ecosystem. As sovereign AI, open innovation, and strategic international collaboration converge, India is transitioning from a fast-growing AI market to a globally competitive AI infrastructure powerhouse.
