National Supercomputing Mission

The National Supercomputing Mission (NSM) was approved by the Union Cabinet of India in March 2015 with a seven-year outlay of ₹4,500 crore, conceived as a joint undertaking of the Department of Science and Technology (DST) and the Ministry of Electronics and Information Technology (MeitY). Implementation responsibility rests with the Centre for Development of Advanced Computing (C-DAC), Pune, and the Indian Institute of Science (IISc), Bengaluru. The mission's legal and administrative basis lies in successive Cabinet financial sanctions rather than a standalone statute; it operates under the broader policy architecture of the Digital India programme and India's longstanding high-performance computing (HPC) tradition, which dates to the development of the first PARAM 8000 by C-DAC in 1991 after the United States restricted the export of Cray supercomputers to India. The NSM was thus framed explicitly as a strategic-autonomy measure, intended to reduce dependence on imported HPC hardware and to embed indigenous design capacity.

The mission proceeds through three escalating phases of indigenisation. In Phase I, systems were largely assembled from imported components to deliver capability quickly. In Phase II, an increasing share of subsystems—including the motherboards, interconnect, and system software stack—were manufactured and integrated within India. In Phase III, the objective is fully indigenous design, including India-designed processors and interconnect. The procurement and deployment model distributes machines across national academic and research institutions—IITs, IISERs, NITs, central universities, and national laboratories—each hosting a system sized to its workload, while a backbone aggregates aggregate national peta-scale capacity. Each deployment couples the hardware grant with a human-resource component: the mission targets training of tens of thousands of HPC-skilled personnel through C-DAC's training programmes and a network of nodal centres.

A central technical thread of the NSM is the development of indigenous building blocks. C-DAC developed the Rudra server platform—an India-designed HPC server board—and the Trinetra high-speed interconnect intended to replace imported fabrics such as InfiniBand. India's parallel processor effort under the Digital India RISC-V (DIR-V) programme, including the Vega and AUM processor families, feeds into the long-term goal of an India-designed CPU at the heart of NSM machines. The systems deployed under the mission carry the PARAM brand—continuing the lineage of C-DAC's 1991 PARAM 8000—and include installations such as PARAM Shivay at IIT-BHU (Varanasi), PARAM Pravega at IISc, PARAM Smriti, PARAM Seva, PARAM Ananta, PARAM Porul, and PARAM Rudra units. Capability is measured in petaflops, with applications spanning computational fluid dynamics, drug discovery, genomics, seismic imaging for oil exploration, disaster simulation, and—through the National Centre for Medium Range Weather Forecasting—weather and climate modelling.

Named contemporary milestones illustrate the mission's trajectory. PARAM Shivay, commissioned at IIT-BHU in 2019, was the first system built fully under the NSM. In September 2024, Prime Minister Narendra Modi dedicated three PARAM Rudra supercomputers—deployed in Pune, Delhi, and Kolkata for research in physics, cosmology, and earth sciences—at a combined cost of approximately ₹130 crore, the systems being notable for their use of indigenous Rudra servers and Trinetra interconnect. The mission's HPC capacity supports the dedicated weather and climate supercomputers Arka and Arunika announced by the Ministry of Earth Sciences in 2024, and feeds the AIRAWAT artificial-intelligence supercomputer hosted at C-DAC Pune, which placed among the world's leading AI systems on the global rankings in 2023.

The NSM should be distinguished from several adjacent initiatives. It is not the IndiaAI Mission, the ₹10,000-plus-crore programme approved in 2024 to build GPU-based AI compute capacity; the NSM concerns general-purpose scientific HPC, though the two increasingly overlap as AI workloads migrate onto HPC infrastructure. It is also distinct from the standalone meteorological supercomputing procurements of the Ministry of Earth Sciences and from the DIR-V chip programme, which supplies processor IP rather than complete systems. Whereas the broader Digital India programme addresses citizen-facing digital services and connectivity, the NSM is a deep-infrastructure and strategic-capability effort aimed at the research and defence-adjacent computing base.

Controversies and edge cases have centred on pace and indigenisation depth. Critics note that the mission's deployment schedule slipped well past its original 2022 horizon, that early-phase systems relied heavily on imported processors and accelerators, and that India's aggregate supercomputing capacity remains modest relative to the United States, China, the European Union, and Japan. The dependence on foreign GPUs—predominantly NVIDIA—for AI-intensive workloads exposes the programme to export-control and supply-chain risk, the very vulnerability the 1991 Cray episode first revealed. The maturation of Rudra, Trinetra, and RISC-V-based processors is the mission's answer, but verifiable fully-indigenous peta-scale systems at scale remain a work in progress.

For the working practitioner, the National Supercomputing Mission is a touchstone case study in technology sovereignty and dual-use industrial policy, and it recurs in the UPSC Civil Services examination under General Studies Paper III's science-and-technology and indigenisation themes. Diplomats and policy analysts track it as an indicator of India's posture on export controls, semiconductor strategy, and its negotiating leverage in technology partnerships such as the India–US iCET initiative. Desk officers handling science-and-technology cooperation should know the C-DAC and IISc implementing roles, the three-phase indigenisation logic, and the PARAM and Rudra nomenclature, because these terms anchor India's claims to an autonomous computing base in bilateral and multilateral fora.