IndiaAI Mission scales up to 34,000+ GPUs, backs Indian AI models for the future

In a major push to scale up its artificial intelligence capabilities, the IndiaAI Mission has empanelled an additional around 16,000 Graphics Processing Units (GPUs) in the second round. With this latest addition, the total number of GPUs available under the mission has now reached 34,333—marking a significant step forward in India’s ambition to create a strong, self-reliant, and inclusive AI ecosystem.

In the second round of empanelment under the IndiaAI Mission, seven Cloud Service Providers (CSPs) have been onboarded to deliver GPU computing power through the IndiaAI Compute Platform, who will be joining existing partners such as Jio Platforms, NxtGen Data Centre, CtrlS Datacenters, Tata Communications, and E2E Networks, in establishing a nationwide distributed AI infrastructure. 

Several Indian firms are powering the IndiaAI Compute initiative with top-tier AI hardware. Cyfuture India, Ishan Infotech, Netmagic, Sify, Vensysco, Locuz, and Yotta are offering a mix of Nvidia GPUs (like H100, H200, B200), AMD chips (MI300X, MI325X), and Intel’s Gaudi series. Vensysco and Locuz also took part in the first round.

Through the IndiaAI Compute Portal, users now have access to a vast pool of AI chips: over 15,000 Nvidia H100 units, along with B200s, H200s, and others. The portal also includes AMD GPUs, AWS Inferentia 2 units, Intel Gaudi chips, and more, bringing scalable, affordable AI compute power to startups, researchers, and enterprises across India.

“In one and a half years, we have made significant progress. As we have completed the second round of GPU empanelment, the third round will also be complete very soon, which will bring another round of GPUs,” said Union Minister, Ashwini Vaishnaw. 

“It's very important that technology should not be left in the hands of a few. A larger section of the society should have access to technology, should be able to develop new solutions, and have better opportunities. That's the philosophy with which the IndiaAI mission was created with seven pillars,” he added. 

Launched in March 2024 with a ₹10,372 crore budget, the IndiaAI Mission aims to build a robust and inclusive AI ecosystem through seven key pillars. It includes a scalable GPU-based compute infrastructure, indigenous foundational model development, access to quality datasets, sector-specific AI applications, upskilling through AI labs, funding for deep-tech startups, and strong safeguards for ethical, responsible AI deployment.

Talking about the IndiaAI Mission, Vaishnaw underscored the Modi government’s holistic approach towards ecosystem development, drawing parallels with the semiconductor sector—where efforts span the entire value chain, from chip fabrication and OSAT to design, critical materials, and manufacturing equipment. He also highlighted the government’s commitment to talent development, with AI-focused skilling initiatives being rolled out across 240 universities nationwide.

Other than the GPU empanelment, the government has shortlisted three more startups for building India's foundation models, for developing and deploying indigenous functional models trained on India specific data. Since the call for proposals launch, over 500 proposals have been received till April 30, 2025. On 26th April, Sarvam AI was announced as the first startup to build India's foundational model. 

Soket AI will build India’s first open-source 120-billion parameter foundation model optimized for linguistic diversity, with applications in defence, healthcare, and education. Meanwhile, Gnani AI is set to develop a 14-billion parameter multilingual Voice AI foundation model capable of real-time speech processing and advanced reasoning.