Yotta and NVIDIA Launch Shakti Cloud on DGX Cloud Lepton to Power India’s Sovereign AI Ambitions

• 20 May 2025
• TM Team
• 0 Comments

Yotta Data Services, a leading sovereign cloud infrastructure and platform services provider, announced that it is teaming with NVIDIA to advance sovereign AI development across India and Southeast Asia by making its GPU compute resources available on NVIDIA’s DGX Cloud Lepton platform.

Yotta’s Shakti Cloud will integrate with DGX Cloud Lepton software and host the training of an LLM designed exclusively for Indian languages and use cases. Developed in partnership with the Government of India under the IndiaAI Mission, this model serves as a foundation for the nation’s generative and agentic AI development.

As part of this integration, Yotta becomes one of the first participants in the new DGX Cloud Lepton marketplace. Yotta is also one of only five Reference Platform NVIDIA Cloud Partners globally and the first NVIDIA Cloud Partner (NCP) in the Asia-Pacific region to join the NVIDIA Exemplar Clouds initiative.

“AI cloud computing is pivitol to India’s digital future,” said Sunil Gupta, Co-founder, CEO & Managing Director, Yotta. “As an empaneled partner in the India AI Mission, Yotta is proud to contribute to the nation’s vision by delivering sovereign, high-performance GPU cloud infrastructure through NVIDIA DGX Cloud Lepton. This enables startups, enterprises, and researchers to build homegrown large language models and AI solutions that reinforce digital sovereignty and innovation at scale.” He further added, “As part of our commitment, in the near future we are deploying the latest NVIDIA B 200 GPUs to support advanced AI workloads, from LLMs and recommender systems to generative applications, accelerating India’s leadership in AI.”

“India stands at the threshold of a new era, where cloud-powered NVIDIA AI infrastructure can serve as an engine of innovation for over a billion people,” said Alexis Bjorlin, vice president of DGX Cloud, NVIDIA. “With Yotta on the DGX Cloud Lepton marketplace, local enterprises and startups can build and deploy world-class AI to empower India’s digital transformation.”

The platform integrates with the NVIDIA software stack, including NVIDIA NIM and NeMo microservices, NVIDIA Blueprints and NVIDIA Cloud Functions (NVCF), to accelerate and simplify the development and deployment of AI applications.

Yotta Integration with DGX Cloud Lepton Software Stack Supercharges India AI

Yotta is adopting DGX Cloud Lepton GPU management software, which delivers real-time GPU health diagnostics and automates root-cause analysis, eliminating manual operations and reducing downtime.

Yotta’s Shakti Cloud GPUs coming to the DGX Cloud Lepton marketplace reside in Yotta’s Tier IV certified NM1 data center, Mumbai and D1, North India’s largest data center, Greater Noida, supporting sovereign AI with data residing within India’s borders and under India’s jurisdiction. The DGX Cloud Lepton software stack enables real-time GPU health monitoring and automated workload orchestration across Yotta’s multi-region infrastructure. Developers can discover and procure on-demand or reserved GPU instances, supporting seamless access from training to deployment.

Sarvam To Train India’s Sovereign LLM With DGX Cloud Lepton

Sarvam will be the first to access Yotta Shakti Cloud with NVIDIA DGX Cloud Lepton and NVIDIA NeMo to train India’s sovereign Large Language Model. Sarvam is on a mission to build the bedrock of Sovereign AI for India and make GenAI a reality for Bharat. The team has already demonstrated proven capability in developing foundational models proficient in Indian languages.

“We’re building multi-modal, multi-scale foundation models that are capable of reasoning, designed for voice, and fluent in Indian languages,” said Vivek Raghavan, Co-founder of Sarvam. “Training our models on Yotta’s Shakti Cloud, powered by NVIDIA DGX Cloud Lepton, is a meaningful step in advancing India’s sovereignty in AI. This partnership will help us bring AI into real-world use, from citizen services to enterprise applications, all built and run within the country.”