‘NVIDIA is AI in India,’ said Jensen Huang at NVIDIA AI Summit 2024 in Mumbai when announcing that Indian AI factories would see 20x NVIDIA GPU growth by year-end.
Citing Yotta, Tata Communciations, and others, Jensen said that NVIDIA in India have a really rich ecosystem in just a span of one year. “By the end of this year, we will have nearly 20 times more compute here in India than just a little over a year ago,” he added.
India’s cloud infrastructure providers and server manufacturers are significantly expanding their data center capacity, with NVIDIA GPU deployment expected to grow nearly tenfold by the end of the year compared to 18 months ago. The announcement was made during the NVIDIA AI Summit in Mumbai.
Tens of thousands of NVIDIA Hopper GPUs will be added to support AI factories—large-scale data centers focused on AI production. These GPUs will provide 180 exaflops of computing power for sectors such as healthcare, financial services, and digital content creation. Leading companies spearheading this initiative include Yotta Data Services, Tata Communications, E2E Networks, and Netweb.
Yotta Data Services will integrate NVIDIA Hopper GPUs into its Shakti Cloud platform, allowing Indian businesses and researchers to adopt generative AI and other AI tools more efficiently.
“The future of AI is about speed, flexibility and scalability, which is why Yotta’s Shakti Cloud platform is designed to eliminate the common barriers that organisations across industries face in AI adoption,” said Sunil Gupta, cofounder, CEO and managing director of Yotta.
Yotta’s customers include Sarvam AI, working on AI models that support major Indian languages; Innoplexus, developing an AI-powered life sciences platform for drug discovery; and Zoho Corporation, creating language models for enterprise use.
Yotta customers using NVIDIA AI Enterprise can tap into NVIDIA NIM, a suite of microservices that enhance AI inference, along with NVIDIA NIM Agent Blueprints, which provide adaptable reference architectures for generative AI projects. This enables quick adoption of advanced AI technologies for tasks such as biomolecular generation, virtual avatar development, and language generation.
Tata Communications will deploy NVIDIA Hopper GPUs across its public cloud infrastructure, expanding its offerings next year to include NVIDIA Blackwell GPUs.
Tata Communications will provide customers with access to NVIDIA AI Enterprise, which includes NVIDIA NIM, NIM Agent Blueprints, and NVIDIA Omniverse—a platform and operating system used by developers to create physical AI and simulate robotic systems.
“By combining NVIDIA’s accelerated computing infrastructure with Tata Communications’AI Studio and global network, we’re creating a future-ready platform that will enable AI transformation across industries,” said A.S. Lakshminarayanan, managing director and CEO of Tata Communications.
E2E Networks is enhancing its cloud infrastructure with NVIDIA Hopper GPUs to support large-scale AI models. Meanwhile, Netweb is increasing its range of Tyrone AI systems to boost India’s AI research and enterprise capabilities.
The company will provide access to clusters featuring NVIDIA Hopper GPUs, interconnected with NVIDIA Quantum-2 InfiniBand networking. This setup will support various applications, including simulations, foundation model training, and real-time AI inference.
Meanwhile, Netweb is expanding its Tyrone AI systems, which are based on NVIDIA MGX, a modular reference architecture designed to accelerate enterprise data center workloads. The new systems will feature NVIDIA GH200 Grace Hopper Superchips, providing the computational power necessary for large hyperscalers, research centers, and supercomputing facilities across India and Asia.
Sanjay Lodha, Chairman and Managing Director of Netweb, said, “Our next-generation systems will help the country’s businesses and researchers build and deploy more complex AI applications trained on proprietary datasets.”
These developments are expected to drive innovation in AI applications, including large language models, scientific visualisations, and industrial AI projects.
The post ‘NVIDIA is AI in India’ appeared first on AIM.
