Abu Dhabi is set to establish a national-scale artificial intelligence (AI) supercomputer in India with a projected computing capacity of 8 exaflops, marking a significant milestone in the expansion of the country’s AI infrastructure.
The announcement was made on the sidelines of the AI Impact Summit 2026 held in New Delhi. The initiative reflects growing technological collaboration between India and the United Arab Emirates, particularly in emerging fields such as artificial intelligence, defence, energy, and space technology.
The development follows the fifth India-UAE Strategic Dialogue conducted in December 2025, as well as the official visit of UAE President Sheikh Mohamed bin Zayed Al Nahyan to India in January 2026, which helped deepen bilateral cooperation across multiple strategic sectors.
The AI supercomputing system will be jointly delivered by G42, a global technology group based in Abu Dhabi, and Cerebras Systems, a United States-based manufacturer known for its high-performance AI computing solutions.
The project will also involve the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), a postgraduate research institution specialising in artificial intelligence, alongside Centre for Development of Advanced Computing (C-DAC), India’s government-run organisation focused on supercomputing and advanced IT research.
An 8-exaflop computing system is capable of performing approximately eight quintillion calculations per second — placing it among the most powerful computing platforms in the world. Officials believe the installation will dramatically enhance India’s domestic capabilities to train, deploy, and scale advanced AI models across sectors including healthcare, education, manufacturing, and smart infrastructure.
The supercomputer will be hosted within India and operate under governance frameworks defined by Indian authorities. According to officials, all data generated and processed through the system will remain within national jurisdiction to ensure compliance with sovereign security standards and regulatory requirements.
The initiative builds upon previous cooperation between G42 and MBZUAI. In December 2025, the two organisations introduced NANDA 87B, an open-source Hindi-English large language model comprising 87 billion parameters. The model was designed to support AI-driven applications in regional languages and improve accessibility to advanced AI technologies across diverse linguistic communities.
The new project signals a broader commitment by both India and the UAE to invest in next-generation computing infrastructure and strengthen collaborative research in artificial intelligence. Experts say such initiatives could accelerate innovation, digital transformation, and technological self-reliance in the coming years.
