🗞️ Why in News A PIB backgrounder of 4 October 2026 sets out the status of the National Supercomputing Mission (NSM): as of September 2026, India has deployed 40 supercomputers with a combined capacity of 68 petaflops (PF), against the mission’s plan of 50 systems and more than 123 PF. It also reports 6,000 indigenous Rudra servers in service and 16,000-plus researchers using the machines. The release precedes ESTIC-2026, the Emerging Science, Technology and Innovation Conclave, to be held at Bharat Mandapam, New Delhi, from 27 to 29 October.
The Mission at a Glance
| Feature | Detail |
|---|---|
| Launched | April 2015 |
| Outlay | About Rs 4,500 crore |
| Steered by | Department of Science and Technology (DST) and Ministry of Electronics and Information Technology (MeitY), jointly |
| Implemented by | C-DAC, Pune and Indian Institute of Science (IISc), Bengaluru |
| Approach | A three-phase “Build approach”: assembly in India, then manufacturing in India, then design and manufacture of key components in India, run concurrently |
| Network backbone | The National Knowledge Network (NKN) links the machines so researchers anywhere can use them |
| Plan | 50 supercomputers, cumulative capacity above 123 PF, across academic and research institutions |
Where It Stands (September 2026)
| Class of system | Number |
|---|---|
| High-end, above 1 PF | 13 |
| Mid-range, 500 TF to 1 PF | 12 |
| Below 500 TF | 15 |
| Total | 40 systems, 68 PF |
The mission has reached 80 per cent of its target in the number of machines but only about 55 per cent in computing power (68 of 123 PF). For scale, the PIB note puts the world’s fastest supercomputer at about 2.19 exaflops in 2026; the whole NSM fleet is about 3 per cent of that single machine.
Measuring speed. Supercomputer power is counted in FLOPS, floating-point operations per second.
| Unit | Operations per second |
|---|---|
| Gigaflop (GF) | 109, one billion |
| Teraflop (TF) | 1012, one trillion |
| Petaflop (PF) | 1015, one quadrillion |
| Exaflop (EF) | 1018, one quintillion |
What India Now Makes Itself
| Component | Status |
|---|---|
| Rudra servers | Designed by C-DAC, technology transferred to Indian electronics manufacturers; 6,000 deployed in PARAM Rudra systems, 1,500 more being manufactured |
| Interconnect | High-speed node-to-node network developed and tested at 100 Gbps and 200 Gbps |
| Cooling | Indigenous cooling technology demonstrated, now being deployed |
| System software | A complete HPC software stack integrated with the machines |
| PARAM Shavak | A “supercomputer in a box” for engineering colleges and universities |
The PARAM lineage. C-DAC’s PARAM 8000, unveiled in 1991, ran at about 1 gigaflop and marked India’s entry into indigenous supercomputing after the denial of imported machines; PARAM Yuva later reached 54 teraflops. In September 2024, the Prime Minister dedicated three PARAM Rudra systems at Pune, Delhi and Kolkata.
What the Machines Are Used For
| Application | Use |
|---|---|
| Flood early warning | Forecasts floods up to two days ahead; in use for the Mahanadi basin |
| Forest-fire spread model | Satellite data plus simulation; tested in the Sikkim Himalayas |
| Urban environment decision support | Predicts heavy rain and air pollution for cities |
| Genomics and drug discovery | Screened existing drugs during COVID-19 and predicted side effects such as heart risk |
| Seismic imaging | Maps subsurface structures for oil and gas exploration |
| Users | IMD, Central Water Commission, CPCB and the Ministry of AYUSH, among others |
People. NSM machines have served more than 16,000 researchers, including 2,900-plus PhD scholars at over 400 institutions, run more than 1.5 crore compute jobs and contributed to over 1,990 research papers. Training runs through hackathons, faculty programmes with AICTE, an NPTEL course on SWAYAM and the HPC Shiksha portal. The government says the mission contributes to 11 Sustainable Development Goals.
Why It Matters
1. Strategic autonomy. India built PARAM because it was denied imported supercomputers in the 1980s. Indigenous servers, interconnects and software reduce that dependence again at a time of export controls on advanced chips.
2. Data and AI. India generates nearly 20 per cent of the world’s data, the PIB note says. Weather, climate, genomics and AI models all need computing at scale; the gap between 68 PF and global exascale machines is the gap Indian researchers work within.
3. The design question. The servers are designed in India, but the processors inside them are still largely imported. The third phase of the Build approach, designing key components, is the hardest and the one that decides self-reliance.
4. Access, not just capacity. Shared machines over the NKN let a researcher in a small university run simulations once possible only at a few national labs.
UPSC Relevance
GS Paper 3. Science and technology: developments and their applications; indigenisation of technology; IT and computers.
A question worth preparing. “India’s National Supercomputing Mission has built capacity but not yet capability.” Assess the mission’s progress and the challenges in achieving self-reliance in high-performance computing. (250 words)
The Mains framing. Start with the 1991 PARAM origin and the 2015 mission; give the numbers (40 systems, 68 PF, 6,000 Rudra servers); show applications in disaster management and health; weigh the gaps (processor dependence, energy use, distance from exascale); end with a way forward: indigenous processors, exascale plans, AI-HPC convergence and wider user training.
📌 Facts Corner, Knowledgepedia
Prelims, statement-ready facts:
- NSM launched April 2015, about Rs 4,500 crore; steered by DST and MeitY; implemented by C-DAC Pune and IISc Bengaluru.
- As of September 2026: 40 supercomputers, 68 PF; plan 50 systems, 123+ PF.
- PARAM 8000 (C-DAC, 1991): about 1 GF, India’s first indigenous supercomputer.
- Rudra: indigenous server; PARAM Shavak: supercomputer in a box; NKN: network backbone.
- 1 PF = 1015 FLOPS; 1 EF = 1018 FLOPS.
- ESTIC-2026: 27 to 29 October, Bharat Mandapam, organised by CSIR.
Prelims, the traps:
- NSM is steered by DST and MeitY together, not by the Department of Space or ISRO.
- C-DAC is a society under MeitY; IISc is the other implementing agency, not IIT Bombay.
Mains, arguments and keywords:
- Strategic autonomy; Build approach; exascale gap; AI-HPC convergence; HPC as a public good; technology denial regimes.
Interview, be ready for:
- “Should India buy foreign supercomputers to close the gap faster?” A mix: buy where time matters, as IMD has, but keep the Build approach alive, because the 1980s denial showed what dependence costs.
Sources: PIB backgrounder, 4 October 2026, Department of Science and Technology, NSM, NSM applications
Source: National Supercomputing Mission: 40 Systems, 68 Petaflops — Ujiyari.com | Free UPSC & State PCS Current Affairs
