National Supercomputing Mission (NSM) & High-Performance Computing In India: Progress And Prospects, 2025

Policy Update
Parth Lathiya

Background

In April 2015, the Government of India launched the National Supercomputing Mission (NSM) as a joint initiative of the Ministry of Electronics and Information Technology (MeitY) and the Department of Science & Technology (DST). The mission has an initial budget of ₹4,500 crore over seven years and set out to create a network of over 70 supercomputers, ranging from 50 teraflops to multi-petaflops systems. These systems are interconnected via the National Knowledge Network (NKN), which connects academic institutions and R&D labs over a high-speed network.

The mission design acknowledges India’s growing computational demands across critical sectors such as climate modelling, materials design, genomics, and artificial intelligence and aims to provide domestic researchers with world-class computational resources to create a robust ecosystem for high-performance computing applications. It also supports the government’s Digital India and Make in India initiatives and eventually helps to reduce India’s reliance on imported systems.

Functioning

The National Supercomputing Mission (NSM) operates through four verticals: Facilities and Infrastructure, Applications, Human Resource Development and Research and Development. Rather than following a conventional top-down approach, the NSM operates through a distributed intelligence model and has created an integrated governance structure (see Table 1) for the effective implementation of programmes.

Table 1: Governance Structure

Vertical 1: Infrastructure Deployment with a Three-Phase “Buy-to-Build” Approach

During the initial phase (2015–18), the mission focused on installing six supercomputers, mainly using imported components, with 30% indigenous input to establish a local assembly line. Later, Phase II (2018-22) emphasised establishing systems with 50% domestic content, a move toward self-reliance. Phase III (2022–25) represents the mission’s full push toward Make-in-India systems which aim to complete a network of more than 70 supercomputers across the country. In this phase, both hardware and software stacks are developed locally. 

Vertical 2: Promotion of Indigenous Research and Development

This vertical aims to reduce India’s dependence on imported supercomputing technology via building critical hardware and software capabilities in-house. To lead this effort, NSM recently brought together research institutions like C-DAC, IISc Bengaluru, and several partner labs to create a collaborative R&D ecosystem which focuses on innovation and self-reliance.

Vertical 3: Applications Development and Support 

The mission also aims to develop high-impact applications that make full use of supercomputing power.  The target was set to create more than 30 domain-specific applications—covering climate science, molecular dynamics, fluid mechanics, AI, and bioinformatics— and specifically optimised to run efficiently on indigenous systems such as PARAM Rudra, using tools like MPI, OpenMP, and GPU libraries. Additionally, each supercomputing site includes high-capacity storage with secure backups, parallel file systems (like Lustre), and encrypted connections through the National Knowledge Network (NKN), which ensure fast and secure data handling for advanced research.

Vertical 4: Human Resource Development (HRD)


The mission’s objective is to prepare a skilled workforce that can effectively use high-performance computing (HPC) in India. To achieve this aim, NSM has established five PARAM Vidya training centres located in Pune, Kharagpur, Chennai, Palakkad, and Goa. These centres offer a range of programmes—from workshops and bootcamps to diploma courses—covering key areas such as parallel programming, cluster management, and domain-specific HPC applications.

NSM has also partnered with AICTE and C-DAC to launch a large-scale faculty development initiative. Through this, around 2,500 faculty members from over 1,000 engineering colleges will receive HPC training, which enables them to pass on these skills to an estimated 50,000 students. The mission has also deployed 50 PARAM Shavak units, easy-to-use “supercomputers in a box” at educational institutions across the country.

Performance and Assessment

As of March 2025, 34 supercomputers have been deployed across 24 institutions, offering a combined capacity of 35 petaflops (see Table 2). These systems are not just operational—they’re actively used. With utilisation ranging from 85% to 95% and uptime above 95%, they provide researchers with fast, reliable access, even for time-sensitive projects. Additionally, the mission has  democratised access to HPC, especially for Tier II and Tier III cities, allowing broader academic participation in frontier-level research.

   Table 2: NSM Performance Metrics & Targets (March 2025)

Advancements in Indigenous Supercomputing Technologies

• The PARAM Rudra series, deployed in Pune, Delhi, and Kolkata, marks a major step in India’s self-reliance in supercomputing as these supercomputers are built using indigenously designed and manufactured HPC servers, known as “Rudra”.
• To improve data communication between computing nodes, C-DAC developed Trinetra, a high-speed interconnect being rolled out in three stages, i.e, Trinetra-POC, Trinetra-A (100 Gbps), and Trinetra-B (200 Gbps).
• Param Shivay at IIT BHU (2019) and Param Pravega, installed at IISc Bengaluru in 2022, delivering 3.3 petaflops of power, is currently the largest academic supercomputer in India.

Image: Param Shivay Supercomputer, IIT BHU

• The government has also launched Project AIRAWAT, which aims to create a shared AI compute platform for academia and industries. In 2023, it ranked 75th in the global Top 500 Supercomputing list at the International Supercomputing Conference (ISC), Germany, which shows that India is part of the global supercomputing race.
• Additionally, NSM’s impact extends well beyond academia. It helps the meteorological department improve monsoon forecasting, accelerated drug discovery during the COVID-19 pandemic, and supported urban planning via heat-zone analysis. Indian start-ups are also using HPC for AI development and product simulations as it has significantly reduced the development time of their products.

Emerging Issues

India’s efforts to build fully indigenous HPC systems depend on its domestic semiconductor capabilities. However, the country still relies on imports for over 90-95% of its chip needs as local fabrication facilities are limited, which eventually caused delays in scaling up indigenous hardware development. To meet NSM’s Phase III targets on time (see Table 2), there is an urgent need to strengthen the India Semiconductor Mission and support private players through long-term investment in chip manufacturing. 

Moreover, supercomputers at the petascale level consume between 5 to 7 megawatts of power, which leads to significant operational costs and environmental concerns for India. In comparison, systems like Europe’s LUMI operate with a Power Usage Effectiveness (PUE) of just 1.1 as it uses renewable energy and recycling waste heat, which makes it sustainable for the future. India can adopt similar measures—such as liquid-cooling systems, solar energy, and waste-heat recovery, especially at large NSM centres. 

While NSM has made good progress on hardware, many scientific applications still run on older code that doesn’t perform well on modern multi-core CPUs, GPUs, or high-speed interconnects. At the same time, retaining skilled and talented professionals is a growing concern , as many move abroad or to private firms offering better pay in comparison to India.

Additionally, Mission appears to be falling short of its original deployment targets for 2025. As shown in Table 2, only 34 supercomputers have been deployed out of the planned 70+, and the total compute capacity has reached only 35 petaflops, still below the 50+ PF target.  At the same time, as NSM platforms begin handling increasingly sensitive data, there is a need to ensure end-to-end encryption, regular security audits, and robust access controls over the National Knowledge Network, which will help to maintain data integrity and public trust.

Way Forward

Looking ahead, the National Supercomputing Mission must accelerate its roadmap through scaling indigenous hardware, deploying AUM processors, and expanding platforms like AIRAWAT. As the Indian government describes NSM as a “first-of-its-kind attempt to boost the country’s computing power,” implementing this vision will require not only stronger local chip manufacturing but also a committed investment in skilled human capital who can carry this technology forward—making India not just a user but a leader in global supercomputing.

References

1. Centre for Development of Advanced Computing (C-DAC). (2023, May 24). India’s AI supercomputer “AIRAWAT” secures 75th rank on global Top500 list. https://www.cdac.in/index.aspx?id=pk_itn_spot1334
2. Centre for Development of Advanced Computing (C-DAC). (n.d.). High performance computing: Technologies and applications. https://www.cdac.in/index.aspx?id=tap&tapcat=High_Performance_Computing
3. Chatterjee, R. (2024, Dec). The semiconductor problem of India: Challenges and future roadmap. Global Scientific Journal, 9(11), 238–248. https://www.globalscientificjournal.com/researchpaper/The_Semiconductor_Problem_of_India.pdf
4. IEEE Spectrum. The race to build the next generation of supercomputers. https://spectrum.ieee.org/nextgeneration-supercomputers
5. LUMI Consortium. (2022). LUMI: Europe’s most powerful supercomputer is solving global challenges and promoting a green transformation. https://www.lumi-supercomputer.eu/lumi-europes-most-powerful-supercomputer-is-solving-global-challenges-and-promoting-a-green-transformation/
6. Ministry of Science and Technology. (n.d.). National Supercomputing Mission | Department of Science and Technology. https://dst.gov.in/national-super-computing-mission
7. Press Information Bureau. (2025, April 28). National Supercomputing Mission. Government of India. https://www.pib.gov.in/PressReleasePage.aspx?PRID=2124920

About the contributor: Parth Lathiya is a Research Intern at the Impact and Policy Research Institute (IMPRI). He is currently pursuing his Master’s degree in Political Science and International Relations at the University of Hyderabad (HCU), Telangana. His academic interests lie in emerging technologies and their regulatory frameworks in India.

Acknowledgement: The author sincerely thanks Ms. Aasthaba Jadeja and IMPRI fellows for theirs valuable contribution.

Disclaimer: All views expressed in the article belong solely to the author and not necessarily to the organisation.

Read more at IMPRI:

Empowering Citizens Digitally: The Role of the National Portal of India

Overseas Workers Research Centre (OWRC), Migrant Resource Centres (MRC)