TLDR: Radian Arc is expanding its GPU capacity in India, aligning with the nation’s aggressive push to significantly increase its high-performance computing resources. This expansion is primarily driven by the escalating demand for AI, machine learning, and other intensive applications, with India’s overall GPU capacity already surpassing 34,000 units and projected to reach 100,000 within the next 18 months.
Radian Arc, a prominent provider of GPU-powered solutions, is actively expanding its high-performance computing infrastructure across India. This strategic move is a direct response to the burgeoning demand for advanced computational resources, particularly for artificial intelligence (AI) and machine learning (ML) applications, and aligns with India’s broader national initiatives to bolster its digital capabilities.
One significant step in this expansion was the strategic partnership forged between Radian Arc and GTPL Hathway Limited in December 2024. This collaboration aims to democratize access to high-performance GPU computing throughout India. David Cook, CEO of Radian Arc, emphasized the importance of this initiative, stating, ‘Our collaboration with GTPL represents a strategic advance in Radian Arc’s mission to democratize access to high-performance GPU computing.’ He added that the partnership would ‘enable organizations throughout India to harness the full potential of GPU-powered solutions for AI innovation and digital transformation, while maintaining strict data sovereignty requirements.’ The initial focus of this partnership is on delivering AI and ML infrastructure services, with plans to extend into areas such as computer vision, academic research, and high-performance computing applications. The platform is designed to support enterprise-grade GPU infrastructure for AI model training, real-time inferencing, large language models, generative AI workloads, and complex data analytics.
This expansion by Radian Arc is set against the backdrop of India’s ambitious national drive to significantly scale its AI computing capacity. According to a Press Information Bureau (PIB) announcement in May 2025, India’s common compute capacity had already surpassed 34,000 GPUs. This figure includes an additional 15,916 GPUs, bringing the total to 34,333 units, supported by various industry partners. Union Minister for Electronics & Information Technology, Shri Ashwini Vaishnaw, highlighted these initiatives at the ‘IndiaAI- Make AI in India, Make AI work for India’ event, stressing that this expanded capacity is ‘crucial to develop indigenous foundational models and AI solutions tailored to the Indian context.’ Minister Vaishnaw also underscored Prime Minister Shri Narendra Modi’s vision for technology democratization, stating, ‘Technology should not be left in the hands of a few. It’s very important that a larger section of the society should be able to access technology.’
The demand for these resources is immense. Sunil Gupta, Co-Founder and CEO of Yotta Data Services, predicted in August 2025 that India’s compute capacity is poised to cross 100,000 GPUs within the next 18 months. He noted that the India AI mission has already aggregated 38,000 GPUs through public-private partnerships, with demand far outstripping supply. Gupta elaborated, ‘If you look at the 500-plus proposals received by the government for training large and small language models, the requirement easily crosses 1–2 lakh GPUs. In my view, 1 lakh GPUs will be operational in the next 18–24 months, possibly sooner.’
Also Read:
- Team Computers and Yotta Partner to Drive Enterprise AI Adoption Across India
- Beyond Prompting: India’s Imperative for Strategic AI Leadership
Further government initiatives include the IndiaAI Compute Portal, launched in February 2025, which offers subsidized computing power to ministries, startups, and researchers. This program is expected to cover 40% of compute costs and provides access to 14,000 GPUs, with 70% being high-end Nvidia H100 models, essential for developing foundational AI models that power advanced chatbots like ChatGPT and Gemini.
