AMD, Ministry of Earth Sciences unveil supercomputers to combat climate challenges

Global chipmaker, AMD, in collaboration with the Ministry of Earth Sciences (MoES) in India, on Thursday launched two high-performance supercomputers—Arka and Arunika—to address climate-related challenges. These computing clusters, named after the Sun, were officially inaugurated by Prime Minister Narendra Modi in September last year.

Arka and Arunika rank among the top three supercomputers in India and stand out as the nation’s most powerful CPU-based systems. Unlike other supercomputers that rely on GPU-based computing for enhanced performance, these systems achieve top-tier efficiency exclusively through CPU processing.

Located at the Indian Institute of Tropical Meteorology (IITM) in Pune, Arka boasts has a computing power of 11.77 PetaFLOPS. Built with 3,021 compute nodes, each integrated with dual AMD EPYC 7643 processors featuring 48-core CPUs, Arka is designed to enhance meteorological research and climate modeling. The system plays a crucial role in improving the accuracy and early prediction of extreme weather events such as tropical cyclones, heavy rainfall, and heatwaves in the Indian subcontinent. 

Meanwhile, Arunika, based at the National Center for Medium Range Weather Forecasting (NCMRWF) in Noida, is focused on production forecasting. With a computational capacity of 8.24 PetaFLOPS and 2,115 compute nodes, this system will aid in studying marine ecosystems and optimizing ocean resource management. Additionally, it will contribute to the development of integrated data models for disaster management across the country.

Jaya Jagadish, Country Head and Senior Vice President of Silicon Design Engineering at AMD India, said, "Our AMD EPYC processors power some of the world’s most efficient and high-performance supercomputers."

Vinay Sinha, Managing Director of Sales at AMD India, stated, "The deployment of these advanced computing clusters marks a major leap in India's weather forecasting abilities. "

The upgraded computational infrastructure will facilitate the creation of advanced models powered by improved technologies such as Artificial Intelligence and Machine Learning. This will enhance last-mile services for various stakeholders by improving the accuracy and efficiency of predictions.