Loading…
The Government set out progress on building technological sovereignty across the AI value chain through the IndiaAI Mission and the India Semiconductor Mission.
The IndiaAI Mission was approved on 7 March 2024 with an outlay of Rs 10,371.92 crore over five years; 20 indigenous foundation model proposals have been selected from 506 applications.
Compute support has reached 237 approved projects and 93.18 lakh sanctioned GPU hours, with a purchase order issued for a roughly 1.1 EFLOPS AI compute system at the NIC Data Centre, Shastri Park, Delhi.
The Union Cabinet approved Semicon 2.0 on 15 July 2026 with an outlay of Rs 1,27,500 crore, building on 12 approved manufacturing projects worth Rs 1.64 lakh crore, three of which are already in commercial production.
India joined the Pax Silica coalition at the India AI Impact Summit 2026 to help secure the global silicon supply chain.
Build an inclusive AI ecosystem covering compute, foundation models, applications, skilling and safe and trusted AI.
Key: Approved 7 March 2024, outlay Rs 10,371.92 crore over five years, implemented by MeitY. Includes 11 national hackathons, 62 AI prototypes developed and 20 AI solutions deployed in public-sector institutions, and 13 selected projects on bias mitigation, machine unlearning, privacy-preserving AI, algorithm auditing and explainability.
Develop a sustainable semiconductor and display ecosystem, covering silicon fabs, display fabs, compound semiconductors, packaging (ATMP/OSAT) and chip design.
Key: Approved by the Union Cabinet on 15 December 2021 with an outlay of Rs 76,000 crore, operationalised through the India Semiconductor Mission under MeitY. 12 projects approved with Rs 1.64 lakh crore committed investment; 3 in commercial production.
Accelerate design and development of Indian-designed chips, expand fabrication units and advanced packaging, and build the wider supply ecosystem.
Key: Approved 15 July 2026 with an outlay of Rs 1,27,500 crore. Notably extends support to materials, gases and equipment - the upstream inputs that Semicon 1.0 did not cover.
Build research and translation capacity in cyber-physical systems through university-anchored hubs.
Key: Implemented by the Department of Science and Technology with an outlay of Rs 3,660 crore. 25 Technology Innovation Hubs established at academic institutions, specialising in AI and machine learning, robotics, IoT, cybersecurity, mining, quantum technologies and fintech.
Extend AI awareness and access to AI resources into universities.
Key: Run under the IndiaAI Mission; a programme was held at Chitkara University, Rajpura, on 6 August 2026.
The AI value chain has four layers, and dependence at any one of them is a vulnerability. At the bottom sit semiconductors - the chips on which everything runs. Above them is compute: the physical GPU clusters and data centres that train and serve models, measured in FLOPS, where an exaflop (EFLOPS) is a quintillion floating-point operations per second. Above that are foundation models - large general-purpose models trained on vast data, which can be adapted to many tasks; a Large Multimodal Model handles text, image, audio and video together, while a Small Language Model is deliberately compact enough to run cheaply or on-device. At the top are applications. India's approach addresses all four simultaneously rather than importing the lower layers: Semicon for chips, the compute empanelment and the Shastri Park system for compute, the 20 selected foundation-model proposals for models, and hackathons and Centres of Excellence for applications. Crucially, intellectual property in the funded models stays with the applicants rather than the Government - the aim is a domestic industry, not a state monopoly.
Simple Analogy: It is the difference between owning a restaurant and owning the farm, the mill and the kitchen too - each layer you do not control is a place someone else can raise the price or close the door.
Implements the IndiaAI Mission and the Semicon India Programme; the nodal ministry for India's AI and semiconductor policy.
The specialised and independent business division within the Digital India Corporation that drives India's semiconductor and display manufacturing strategy and administers the Semicon India Programme.
Implements NM-ICPS and its 25 Technology Innovation Hubs.
The Government's IT services organisation, hosting the roughly 1.1 EFLOPS AI compute system at its Shastri Park Data Centre in Delhi.
An international coalition India joined at the India AI Impact Summit 2026, alongside the United States and other partners, to secure the global silicon supply chain and build a resilient technology ecosystem.
GS Paper 3 > Science and Technology: Developments and Applications, Indigenisation of Technology; Economy: Industrial Policy
General Awareness > Science and Technology, Government Schemes
General Awareness > Technology and Economy
Semiconductor and AI missions have appeared in UPSC Prelims and Mains every year since 2022 and are staple Banking and SSC general-awareness material.
A large general-purpose model trained on broad data and adapted to many downstream tasks.
A model handling multiple input types - text, image, audio, video - together; twelve of the twenty selected proposals are LMMs.
A deliberately compact language model, cheaper to run and deployable on limited hardware; eight of the twenty selected proposals are SLMs.
Exa-FLOPS - a quintillion floating-point operations per second; the ordered NIC system is about 1.1 EFLOPS.
Assembly, Testing, Marking and Packaging / Outsourced Semiconductor Assembly and Test - the back-end stages of chipmaking supported under Semicon India.
An indigenous multilingual foundation-model effort whose released models include Param2-17B, Patram-7B and Shrutam-2.
A system integrating computation, networking and physical processes - the focus of NM-ICPS and its 25 Technology Innovation Hubs.