NEW DELHI – Abu Dhabi will set up a national AI supercomputer with 8 exaflops of computing power in India, ushering in a new phase in India’s AI infrastructure development.
The system will be provided by G42, a global technology group based in Abu Dhabi, and Cerebras, in partnership with Mohammed Bin Zayed University of Artificial Intelligence (MBZUAI) and India’s Center for Advanced Computing Development (C-DAC) to develop the world’s fastest AI inference.
This groundbreaking project was announced on the sidelines of the AI Impact Summit 2026 in New Delhi, India. This follows the 5th India-UAE Strategic Dialogue held in December 2025 and President His Highness Sheikh Mohammed bin Zayed Al Nahyan’s visit to India in January 2026, which solidified the framework for a comprehensive partnership across defence, technology, space and energy.
At 8 exaflops, this new system significantly increases peak computing power and marks India’s move towards exaflops-scale AI infrastructure, expanding the country’s computing capacity for advanced AI development.
The system is hosted within India and operates under an India-defined governance framework, with all data remaining within the country’s jurisdiction. This supercomputer is designed to meet sovereign security and compliance requirements and will serve as a foundational asset for the Indian AI mission.
“Sovereign AI infrastructure is becoming essential to national competitiveness,” said Manu Jain, CEO of G42 India. “This project will bring that capability to India on a national scale, enabling local researchers, innovators and businesses to become AI native while maintaining full data sovereignty and security.”
Once operational, India’s supercomputer will be accessible to the country’s diverse ecosystem, from major institutions to start-ups, small and medium-sized businesses, and government ministries. This democratized access model is designed to lower barriers to AI innovation, especially in applications that serve India’s 1.4 billion people.
“MBZUAI is committed to advancing AI research and education that addresses real-world challenges,” said Richard Morton, executive director of MBZUAI’s Basic Model Institute. “This collaboration with India represents a shared commitment to expand access to advanced AI computing for researchers and students, enabling breakthroughs in critical areas such as medicine, agriculture, and education.”
“Cerebras and G42 have already successfully delivered a Condor Galaxy supercomputer in the U.S., demonstrating that our technology is purpose-built for the most demanding AI workloads at scale,” said Andy Hock, chief strategy officer at Cerebras. “Bringing this system to India represents a significant step forward in India’s computing power and sovereign AI journey. It will accelerate the training and inference of large-scale models and enable researchers and developers to build AI tailored to India’s needs.”
This latest project builds on the G42’s commitment to help countries build their domestic AI capacity. In December 2025, G42 and MBZUAI released NANDA 87B, the latest version of the open source Hindi and English Large Language Model (LLM) with 87 billion parameters.
As one of the world’s fastest-growing digital economies, India plays a central role in driving AI innovation in the region. The new supercomputer aims to strengthen India’s ability to safely build, deploy and scale AI within the country.
