National Supercomputing Mission: Building India’s Self-Reliant HPC Ecosystem

Yogeshwar Sonawane of C-DAC presents India's National Supercomputing Mission (NSM) and its role in building a self-reliant HPC ecosystem. India's HPC journey began in the late 1980s, with C-DAC delivering the country's first supercomputer in 1991; the NSM, launched around 2016-17, rests on four pillars: advanced supercomputing infrastructure, applications of national importance, R&D for exascale readiness including indigenous component development, and trained manpower. To date roughly 39 systems totaling 267 petaflops have been commissioned across national labs and academic institutions, with nine more systems adding 55 petaflops expected shortly; the infrastructure serves over 16,000 users across 50 institutes at about 85 percent utilization, and more than 30,000 people have been trained. Applications include an operational flood forecasting model for the Mahanadi river in use since 2023, seismic imaging software used by ONGC for oil and gas exploration, weather and urban environment modeling, bioinformatics tools for drug discovery and repurposing used by ministries and pharma industry, contributions to India's COVID vaccine development, and AI for India's 22-plus official languages. Indigenous development features the Rudra server platform, designed and manufactured in India and underlying eight deployed systems totaling about 25 petaflops, the Trinetra HPC interconnect in 100G and 200G variants with 400G in progress, a unified system software stack with monitoring tools, and Arm and RISC-V processor design initiatives. NSM 2.0 targets exascale computing around 2030.