Zoho-backed Netrasemi, incubated at Kerala’s TrEST Park, has raised ₹125 crore and launched its first chip for drones, cameras and robotics.
Indian semiconductor ecosystem: News, Updates & Discussions.
- Thread starter Ashwin
- Start date
Zoho-backed Netrasemi, incubated at Kerala’s TrEST Park, has raised ₹125 crore and launched its first chip for drones, cameras and robotics.
TDB-DST signs agreement with BigEndian Semiconductors for ₹130 crore RDI support to develop indigenous AI Vision SoC
Project VeerAI to advance secure, application-specific semiconductor technology from TRL-5 to TRL-9
Posted On: 05 OCT 2026 12:23PM by PIB DelhiStrengthening India’s efforts to build indigenous semiconductor capabilities for critical and emerging applications, the Technology Development Board (TDB), Department of Science & Technology (DST), Government of India, has signed an agreement with M/s BigEndian Semiconductors, Bengaluru, Karnataka for financial support of ₹130 crore under the Research Development and Innovation (RDI) Fund.
The project, Project VeerAI, has a total project cost of ₹260 crore and aims to take the technology from Technology Readiness Level (TRL)-5 to TRL-9. The financial support will be provided through Optional Convertible Debt (OCD). Project VeerAI seeks to develop a camera-focused AI Vision System-on-Chip (SoC) on a sovereign technology platform, addressing the growing requirement for secure, intelligent and application-specific semiconductor solutions.
The project is positioned around the development of a secure AI chip for surveillance applications, particularly CCTV systems, with the objective of reducing dependence on externally developed semiconductor and technology supply chains. The proposed platform is designed to combine AI processing and smart features with security, privacy and connectivity capabilities. The technology is being developed initially for CCTV and surveillance applications in India and the Global South, with a roadmap for deployment across multiple sectors. Beyond surveillance, the platform is envisaged for applications in defence, automotive, industrial and medical domains, creating a wider market for indigenous AI-enabled semiconductor solutions.
The project also seeks to leverage India’s growing VLSI design and semiconductor talent, translating domestic design capabilities into a scalable product platform. By focusing on application-specific SoCs rather than general-purpose chips, the programme aims to address specific industry requirements while enabling faster technology deployment and commercialisation. The development of indigenous AI and semiconductor technologies has growing significance as connected devices and intelligent systems become increasingly embedded across critical infrastructure. Secure and application-specific chips can play an important role in addressing requirements related to data security, privacy, device-level intelligence and technology sovereignty.
Through the RDI support, BigEndian Semiconductors will undertake further technology development, validation, integration, testing and scale-up activities required to progress Project VeerAI from TRL-5 to TRL-9.
QNu Labs, BISAG-N and IIT Gandhinagar Demonstrate 5 km-Scale Free-Space QKD Link
Matt Swayne
October 5, 2026
Insider Brief
- QNu Labs, BISAG-N and IIT Gandhinagar completed a 5.56-kilometer free-space quantum key distribution trial, integrating keys transmitted through the atmosphere with a post-quantum cryptography application.
- The link achieved a secure key rate of 230–260 bits per second with a quantum bit error rate below 5%, supporting successful encryption, transmission and decryption of a test message.
- The hybrid system combined quantum key distribution with post-quantum cryptography and quantum random number generation, with the software security layer designed to remain operational if the quantum channel becomes unavailable.
Conducted on the night of September 27–28, 2026, the trial established a 5.56 km free-space quantum communication link between BISAG-N and IIT Gandhinagar, with QNu Labs’ Pointing, Acquisition and Tracking (PAT) system maintaining the 1550 nm quantum channel. Using the DPS-Decoy QKD protocol, the link achieved a stable QBER below 5% and a secure key rate of 230–260 bps. The trial also demonstrated the end-to-end integration of QKD with BISAG-N’s Vedic Kavach PQC enabled browser and web server, with 256-bit keys delivered through the ETSI GS QKD 014 interface, enabling a test message to be encrypted, transmitted and successfully decrypted.
The trial combined two complementary layers of quantum security. QNu Labs’ QKD technology provided hardware-based quantum key distribution over the free-space optical channel, while Vedic Kavach provided a software layer based on post-quantum cryptography and quantum random number generation (QRNG). The architecture was designed so that the application remains protected even if the physical quantum channel is temporarily unavailable, providing a practical approach to building quantum-resilient communication systems.
Sunil Gupta, Co-founder and CEO, QNu Labs, said, “Quantum security will become meaningful at scale only when technologies can move beyond controlled laboratory environments and operate across real communication infrastructure. This field trial shows that QKD, supported by precise optical pointing and tracking, can be deployed over a multi-kilometer free-space link and integrated with a post-quantum security layer. The successful 5.56 km demonstration is an important step towards developing longer-distance quantum-secure networks and exploring architectures that can eventually support multiple nodes and satellite-based quantum communication.”
The technical performance of the trial was enabled by QNu Labs’ PAT system, which used coarse pointing through a gimbal and fine steering at a 100 Hz tracking bandwidth to maintain alignment between the two terminals. The system achieved coarse pointing accuracy of 5 microradians and fine pointing accuracy of up to 0.1 microradians, enabling the quantum channel to remain aligned despite atmospheric conditions and physical movement along the optical path.
Cmde Manish Tripathi (Retd.), In-charge, ISTF, IIT-G said, “The successful demonstration of a 5.56 km free-space QKD link provides a valuable field environment for advancing practical quantum communication technologies. The collaboration brings together academic research, indigenous technology and real-world experimentation, and demonstrates how these capabilities can be integrated to address the emerging requirements of secure communications.”
Dr. Vinay Thakur, DG BISAG-N said, “The integration of Vedic Kavach with QNu Labs’ QKD technology demonstrates the potential of combining post-quantum cryptography with quantum key distribution as complementary layers of security. The successful end-to-end application test is an important step towards developing practical quantum-resilient communication architectures that can evolve with emerging security requirements.”
At a secure key rate of 230–260 bps, the link generated approximately one fresh 256 bit key every second while maintaining QBER below 5%, demonstrating the viability of a hybrid QKD and PQC architecture in a field environment. The trial provides a foundation for advancing towards longer links, multiple nodes and, eventually, satellite-based quantum communication, while strengthening the resilience of critical communication infrastructure against emerging quantum era security threats.
QNu Labs, BISAG-N and IIT Gandhinagar Demonstrate 5 km-Scale Free-Space QKD Link
DRDO tests quantum-secured communication system on operational military network
System tested at Odisha’s Chandipur range to secure real-time telemetry and critical data during artillery and armoured weapon trials
Vijay Mohan
Tribune News Service
Chandigarh, Updated At : 12:50 PM Oct 14, 2026 IST
Advancing India’s capabilities in quantum-secured communications, the Defence Research and Development Organisation (DRDO) successfully configured and tested a Quantum Key Distribution (QKD) platform on a real, operational military communication network at one of its test ranges.
This demonstrates the use of QKD to fortify real-time critical observations and telemetry associated with artillery and armoured weapon tests and evaluation, according to information shared by the DRDO on Wednesday.
The test was carried out at its Proof and Experimental Establishment (PXE) at Chandipur in Odisha.
“Quantum Key Distribution has been designed and developed under the Microelectronic Devices, Computational Systems and Cyber Security (MCC) cluster of DRDO, with production partner Taqbit labs, and will aptly provide unconditional security along with eavesdropping detection,” the DRDO said.
The ranges at PXE are among those where India tests artillery and armoured platforms as well as some other weapon systems under actual operational conditions.
During such tests, large amounts of critical real-time data, such as telemetry, sensor readings, performance parameters and technical observations, are transmitted between different centres. Protecting that data with QKD implies that not only is the information secure even if there are attempts to intercept it, but also that any interception attempt is spotted immediately, DRDO sources said.
In July this year, the DRDO, in collaboration with its production partner, had completed military field trials of a scalable and practically secure, fibre-based QKD system, thereby moving ahead in providing safe and reliable strategic communication systems.
Quantum communication provides fundamentally unbreakable encryption, making it a dual-use technology with applications in securing data in strategic sectors such as defence, finance and telecommunications, as well as in protecting national security-related communications.
“This successfully lays the foundation for large-scale, multi-hop quantum networks that is indigenously designed and can seamlessly safeguard national communication infrastructure against emerging cyber threats in the quantum era,” the DRDO had then said.
Last year, the DRDO had successfully demonstrated an experimental advancement in quantum communication through the DRDO-Industry-Academia Centre of Excellence (DIA-CoE) at IIT-Delhi. This secure communication paved the way for real-time applications in quantum cybersecurity, including QKD, the development of quantum networks and the future quantum internet.
According to official statements, quantum entanglement-based QKD offers significant advantages over the traditional prepare-and-measure method by enhancing both security and functionality.
Even if communication devices are compromised or imperfect, the use of quantum entanglement ensures secure key distribution. Any attempt to measure or intercept the entangled photons disturbs their quantum state, allowing authorised users to detect the presence of an eavesdropper.
In May 2025, the DRDO had set up a new Quantum Research Centre in Delhi to strengthen indigenous capabilities in quantum computing for strategic and defence applications.
The facility is equipped with state-of-the-art experimental set-ups for research and development in critical quantum domains, including characterisation of experimental platforms for developing and validating quantum key distribution techniques to enable ultra-secure communication and safeguard national security in the post-quantum era.
DRDO tests quantum-secured communication system on operational military network - The Tribune
System tested at Odisha’s Chandipur range to secure real-time telemetry and critical data during artillery and armoured weapon trials
Vijay Mohan
Tribune News Service
Chandigarh, Updated At : 12:50 PM Oct 14, 2026 IST
During such tests, large amounts of critical real-time data, such as telemetry, sensor readings, performance parameters and technical observations, are transmitted between different centres. Representative image/DRDO
Advancing India’s capabilities in quantum-secured communications, the Defence Research and Development Organisation (DRDO) successfully configured and tested a Quantum Key Distribution (QKD) platform on a real, operational military communication network at one of its test ranges.
This demonstrates the use of QKD to fortify real-time critical observations and telemetry associated with artillery and armoured weapon tests and evaluation, according to information shared by the DRDO on Wednesday.
The test was carried out at its Proof and Experimental Establishment (PXE) at Chandipur in Odisha.
“Quantum Key Distribution has been designed and developed under the Microelectronic Devices, Computational Systems and Cyber Security (MCC) cluster of DRDO, with production partner Taqbit labs, and will aptly provide unconditional security along with eavesdropping detection,” the DRDO said.
The ranges at PXE are among those where India tests artillery and armoured platforms as well as some other weapon systems under actual operational conditions.
During such tests, large amounts of critical real-time data, such as telemetry, sensor readings, performance parameters and technical observations, are transmitted between different centres. Protecting that data with QKD implies that not only is the information secure even if there are attempts to intercept it, but also that any interception attempt is spotted immediately, DRDO sources said.
In July this year, the DRDO, in collaboration with its production partner, had completed military field trials of a scalable and practically secure, fibre-based QKD system, thereby moving ahead in providing safe and reliable strategic communication systems.
Quantum communication provides fundamentally unbreakable encryption, making it a dual-use technology with applications in securing data in strategic sectors such as defence, finance and telecommunications, as well as in protecting national security-related communications.
“This successfully lays the foundation for large-scale, multi-hop quantum networks that is indigenously designed and can seamlessly safeguard national communication infrastructure against emerging cyber threats in the quantum era,” the DRDO had then said.
Last year, the DRDO had successfully demonstrated an experimental advancement in quantum communication through the DRDO-Industry-Academia Centre of Excellence (DIA-CoE) at IIT-Delhi. This secure communication paved the way for real-time applications in quantum cybersecurity, including QKD, the development of quantum networks and the future quantum internet.
According to official statements, quantum entanglement-based QKD offers significant advantages over the traditional prepare-and-measure method by enhancing both security and functionality.
Even if communication devices are compromised or imperfect, the use of quantum entanglement ensures secure key distribution. Any attempt to measure or intercept the entangled photons disturbs their quantum state, allowing authorised users to detect the presence of an eavesdropper.
In May 2025, the DRDO had set up a new Quantum Research Centre in Delhi to strengthen indigenous capabilities in quantum computing for strategic and defence applications.
The facility is equipped with state-of-the-art experimental set-ups for research and development in critical quantum domains, including characterisation of experimental platforms for developing and validating quantum key distribution techniques to enable ultra-secure communication and safeguard national security in the post-quantum era.
DRDO tests quantum-secured communication system on operational military network - The Tribune
QKD is getting operationalized for defence use cases. Let's hope photonics can be operationalized soon.
India plans indigenous GPU chips in 24 months, says MeitY Secretary S Krishnan on AI strategy
MeitY Secretary S Krishnan outlines India's plans for indigenous GPU chips, Semicon 2.0, sovereign AI, data protection and new opportunities for the IT industry amid AI-driven transformation.
By Anoop Verma
ET Government
Updated On Nov 04, 2026 at 07:28 AM IST
India is moving towards developing indigenous GPU chips, expanding domestic semiconductor manufacturing and strengthening its capabilities across the artificial intelligence technology stack.
The Ministry of Electronics and Information Technology (MeitY) is targeting the development of an Indian GPU chip over approximately 24 months, while India Semiconductor Mission 2.0 seeks to build a comprehensive ecosystem encompassing chip design, fabrication, advanced packaging, materials, equipment and research.
In a conversation with Anoop Verma, S. Krishnan, Secretary, MeitY, outlines the government's strategy for reducing dependence on foreign technologies, commercialising indigenous processors such as SHAKTI and VEGA, and supporting the development of Indian AI models.
He also discusses the DEPA framework for secure data sharing, preparations for implementing the Digital Personal Data Protection Act, and the opportunities emerging from AI adoption across enterprises. Drawing a parallel with the Y2K era, Krishnan argues that artificial intelligence can create a new phase of growth for India's IT services industry, generating employment opportunities even as automation transforms existing jobs.
Edited excerpts:
On October 31, MeitY launched the DEPA for AI Training Handbook, outlining a framework for trusted data collaboration while protecting privacy and institutional control. How will this framework be implemented across government departments, research institutions and private enterprises? What mechanisms will ensure that Indian AI developers can access high-quality datasets without compromising citizens' data rights?
The DEPA mechanism is primarily about creating a secure framework through which data can be shared for research, artificial intelligence and other purposes without compromising privacy. We have already created AIKosh, which is operational and provides access to datasets for AI development. However, some datasets contain personally identifiable information, and even when such information is anonymised, there remains a possibility that anonymisation could be reversed.
This raises an important question: why should an entire dataset be supplied to a researcher when the requirement is only to undertake a particular analysis? We are therefore looking at a framework in which researchers can access and analyse data within a secure environment for a specified purpose. They can conduct their research, obtain their findings and take those findings away, but the underlying data remains within that protected environment.
The principle is similar to the reference sections of traditional libraries. There were books that could not be borrowed. Researchers could consult them, make notes and leave, but the material itself remained within the library. We are applying a comparable principle to digital datasets.
The National Health Authority is also adopting a similar approach for medical data, where protecting personal information is particularly important. Our objective is to establish good practices that can be adopted more widely by government agencies, private companies and research institutions. The framework could also be adopted internationally. It provides a way to facilitate data sharing and AI research while preserving privacy, anonymity and the conditions under which the data is made available. This is essential for enabling AI development without compromising citizens' rights.
India continues to depend substantially on imported GPUs and foreign technology providers for advanced AI computing. What is MeitY's long-term strategy for reducing these dependencies? Is India considering indigenous AI accelerators, domestic computing architectures and dedicated sovereign AI infrastructure for sensitive government and strategic applications?
The infrastructure required for artificial intelligence cannot be created overnight. We are taking a studied and comprehensive approach because the AI technology stack consists of five distinct layers, each of which needs to be addressed. The first layer is energy and power infrastructure. India is relatively well placed in this area, which is one reason why several international companies are interested in establishing data centres and AI computing facilities in the country.
The second layer is the chip itself, particularly the accelerator chips used for AI computing. Today, these chips are produced by a relatively small number of companies globally, and that concentration contributes to their high cost. Under India Semiconductor Mission 2.0, we intend to target the development of an indigenous GPU chip. The minister has also indicated this priority. We should have a GPU chip of our own. This will not happen overnight. It is a project that will require approximately 24 months, and we intend to work towards that objective.
The third layer is computing infrastructure, including servers and associated hardware. India already has companies capable of manufacturing servers. Companies such as VVDN, Kaynes and Avalon, among others, have developed manufacturing capabilities in this area. Several server components are also being manufactured domestically, supported by the Electronics Component Manufacturing Scheme and other initiatives.
There are still some components, particularly semiconductor chips, that are not manufactured entirely in India. Addressing those gaps is part of our strategy. Nevertheless, the capability to assemble and manufacture Indian servers already exists. C-DAC has developed its Rudra series of servers, including the AI-capable Rudra 3 series. These systems are increasingly being deployed, and technology transfer to Indian companies is helping expand domestic server manufacturing capabilities. We are also encouraging international companies establishing computing infrastructure in India to develop manufacturing facilities here.
The fourth layer consists of AI models. Under the IndiaAI Mission, we have supported more than 20 indigenous models. We hope that at least two of these will develop into frontier-level models, and we are continuing to support their development. At the same time, we recognise that large general-purpose models are not the only requirement. Smaller, specialised models can be highly effective in sectors such as agriculture, manufacturing, healthcare and education. These models can perform specific tasks efficiently while consuming less energy and computing power.
The fifth layer, and perhaps the most important from India's perspective, is applications. The real benefits of AI will emerge when it is deployed across businesses, industries and public services. Enterprise-level AI adoption remains limited. Individuals may be using AI tools, but many organisations have not integrated AI across their operations. Concerns about security, confidentiality and the possibility of sensitive business data being shared externally are among the reasons.
India has a significant opportunity to address these challenges and support enterprise-level AI adoption, not only domestically but internationally. By developing capabilities across all five layers, we can strengthen our technological resilience and eventually achieve greater strategic autonomy.
The government has approved Semicon 2.0 with an outlay of ₹1,27,500 crore, and notifications covering its six pillars were issued in August. How will this programme move India beyond semiconductor assembly and packaging towards advanced fabrication, indigenous chip design, equipment manufacturing and materials production? What measurable manufacturing milestones should India achieve over the next three to five years?
What we are attempting through India Semiconductor Mission 2.0 is to build the entire semiconductor ecosystem in the country. The first phase of the mission focused primarily on two areas: semiconductor wafer fabrication and the assembly and packaging of chips. We now recognise that India must develop capabilities across the entire value chain, beginning with chip design and extending to manufacturing equipment, materials, chemicals, advanced packaging and research and development.
Chip design is an important part of this strategy. Under the first phase, we introduced the Design Linked Incentive scheme, which has produced encouraging results. We identified more than 100 companies, of which 24 received assistance under the scheme. Subsequently, 15 of these 24 companies also attracted venture capital or other private investment. This demonstrates that government support can help companies develop technologies that subsequently attract commercial investment.
Under ISM 2.0, we intend to place greater emphasis on chip design, including strategically important chips such as AI accelerators. We are planning a higher allocation for this segment. We are also looking at different categories of chip design, including strategic chips and commercial chips that require support. The approach will include investment-based assistance rather than relying exclusively on grants. Support for electronic design automation tools will also continue.
The second major priority is developing the ecosystem of materials, machinery and equipment required for semiconductor manufacturing. A fabrication facility needs an extensive network of suppliers providing specialised gases, chemicals, materials and equipment. These supporting industries are essential for establishing a sustainable semiconductor manufacturing base. Under ISM 2.0, we intend to support their development. Several companies have already expressed interest and announced investment plans in these areas.
The third priority is to differentiate the support provided to various manufacturing activities. Semiconductor fabrication facilities will continue to receive substantial assistance, with support of 40%. Advanced packaging facilities will be eligible for assistance of 35%. The objective of providing higher support for advanced packaging is to encourage the industry to move towards more sophisticated technologies. Under the first phase, we attracted approximately eight or nine companies in conventional packaging. We now want the industry to progress to the next level.
Research and development will be another important component of the programme. We have established capabilities at the 28-nanometre technology node, but progressing towards more advanced nodes involves considerably higher costs. We therefore need to examine how research and technological development can help India move beyond 28 nanometres. This is an important area for future investment.
There are also emerging technologies, including display technologies and silicon photonics, where the government has supported research in various institutions. Some of these initiatives have produced promising results, and we now want to take them further towards practical and commercial outcomes. The overall objective of ISM 2.0 is to ensure that India develops a more complete and sustainable semiconductor ecosystem rather than remaining dependent on a few isolated segments of the manufacturing process.
Recent initiatives under the Chips to Startup programme and the Digital India RISC-V ecosystem have demonstrated growing domestic chip-design capabilities. What is the roadmap for commercialising indigenous processors such as SHAKTI and VEGA? Can India establish a significant market for domestically designed chips in government systems, defence, telecommunications, automotive electronics and industrial applications?
Both SHAKTI and VEGA processors have been developed indigenously using the RISC-V architecture. These developments demonstrate India's growing capabilities in processor design. As I mentioned, 24 companies received assistance under the Design Linked Incentive programme, and 15 of them subsequently raised venture capital funding. These companies are developing a range of products, including microcontrollers, flight controllers for drones, chips for smart meters and chips for CCTV cameras.
These are practical applications for which there is substantial demand. I believe the prospects for commercialising many of these products and taking them to a significantly higher level are strong. The government will continue to support this process. India also has a very large semiconductor market. Its current size is a little over $50 billion, and it is expected to approach $100 billion by 2030. The combined production from the semiconductor projects currently being developed will represent only a fraction of India's total consumption.
However, semiconductor chips serve different purposes, and there cannot be a single approach to developing the market. Different applications require different types of chips, with varying performance, reliability and security characteristics. We need to ensure that Indian chip-design companies have access to an adequate domestic market. That does not necessarily mean imposing restrictions across the entire consumer electronics market, because doing so could create unnecessary burdens.
There are, however, specific areas where the use of trusted and verified chips is particularly important from a security perspective. For example, CCTV cameras, industrial sensors, Internet of Things devices, drone flight controllers and smart meters can involve sensitive information. In such applications, data leakage or a compromise of system integrity can create serious security concerns.
It is therefore reasonable to ensure that the chips used in these systems come from trusted and verified sources. Indian companies could be encouraged to supply chips for such applications, provided their products meet the required standards of quality and performance. This approach can help establish a market for India-designed and India-made chips while also addressing important security requirements. It creates an opportunity for domestic semiconductor companies to develop commercially viable products without imposing unnecessary restrictions on the broader market.
The Digital Personal Data Protection framework is becoming increasingly important for government departments, digital platforms and private enterprises. How is MeitY preparing these organisations for compliance? What are the major challenges in implementing consent management, grievance redressal and regulatory requirements, particularly for smaller businesses and government departments?
We have provided an implementation period of 18 months, and the framework is intended to become operational by May 2027. The government had initially considered bringing forward the implementation date. However, several businesses represented that they would require more time to prepare for compliance. Taking these concerns into account, we have retained the implementation timeline.
We have stepped up our awareness campaign around the Digital Personal Data Protection Act, its provisions and the measures that organisations must take to comply with it. A number of awareness programmes are being conducted in Delhi and other centres across the country. We are also using social media, other communication channels and institutional mechanisms to ensure that businesses, government departments and other stakeholders understand their obligations.
This effort has already helped increase awareness of the legislation and the changes that organisations will need to introduce. The National Informatics Centre has also developed a product to assist government departments and other organisations handling substantial amounts of personal data in complying with the provisions of the Act. This tool is being widely publicised among government ministries.
The Cabinet Secretary has written to all secretaries in the Union government and the chief secretaries of states, emphasising the need to prioritise compliance with the DPDP Act. MeitY is supporting ministries and other government organisations in preparing for implementation. The objective is to ensure that institutions handling personal data are ready to meet their obligations when the relevant provisions become applicable.
At MeitY's June consultation on AI transformation of the IT services industry, you emphasised reskilling and technological upgrading rather than focusing exclusively on potential job losses. With generative and agentic AI transforming software development and business services, how should India's IT industry adapt its business model? What government-industry initiatives are being developed to protect competitiveness, create higher-value employment and prepare the workforce for AI-driven disruption?
NITI Aayog produced a paper last year examining the implications of AI for employment in India's IT and IT-enabled services sector. The paper pointed out that the sector currently employs approximately six million people. It anticipated that around two million existing jobs could become redundant because of technological changes, but that approximately four million additional jobs could be created. This would result in a net increase in employment.
The important question is how these new employment opportunities will emerge. I believe the principal opportunity lies in the deployment of AI across enterprises and organisations. This is an area where India has a substantial advantage. There is a parallel with the Y2K period, which created significant opportunities for the Indian IT industry. At that time, Indian companies were able to provide technology services to businesses around the world, and this contributed substantially to the growth of the sector.
AI could create a similar opportunity, although the nature of the work will be different. It is relatively easy to say that enterprises should adopt artificial intelligence. The actual process is much more complex. AI solutions must be designed around the specific requirements of each organisation. They must protect sensitive data, integrate with existing systems and be deployed effectively.
These are precisely the areas where India's IT industry can make an important contribution. We have a large pool of trained IT professionals with experience in developing, integrating and managing complex technology systems for organisations across the world. These capabilities can be applied to enterprise AI adoption on a substantial scale.
Indian companies can help businesses move from experimenting with AI tools to deploying AI throughout their operations. This will require technological adaptation and new skills, but it also presents a significant commercial opportunity. Our trained IT workforce gives India the capacity to undertake this work at scale and competitively. I believe this is where a major opportunity lies for the next phase of growth of India's IT services industry.
India plans indigenous GPU chips in 24 months, says MeitY Secretary S Krishnan on AI strategy
MeitY Secretary S Krishnan outlines India's plans for indigenous GPU chips, Semicon 2.0, sovereign AI, data protection and new opportunities for the IT industry amid AI-driven transformation.
By Anoop Verma
ET Government
Updated On Nov 04, 2026 at 07:28 AM IST
India is moving towards developing indigenous GPU chips, expanding domestic semiconductor manufacturing and strengthening its capabilities across the artificial intelligence technology stack.
The Ministry of Electronics and Information Technology (MeitY) is targeting the development of an Indian GPU chip over approximately 24 months, while India Semiconductor Mission 2.0 seeks to build a comprehensive ecosystem encompassing chip design, fabrication, advanced packaging, materials, equipment and research.
In a conversation with Anoop Verma, S. Krishnan, Secretary, MeitY, outlines the government's strategy for reducing dependence on foreign technologies, commercialising indigenous processors such as SHAKTI and VEGA, and supporting the development of Indian AI models.
He also discusses the DEPA framework for secure data sharing, preparations for implementing the Digital Personal Data Protection Act, and the opportunities emerging from AI adoption across enterprises. Drawing a parallel with the Y2K era, Krishnan argues that artificial intelligence can create a new phase of growth for India's IT services industry, generating employment opportunities even as automation transforms existing jobs.
Edited excerpts:
On October 31, MeitY launched the DEPA for AI Training Handbook, outlining a framework for trusted data collaboration while protecting privacy and institutional control. How will this framework be implemented across government departments, research institutions and private enterprises? What mechanisms will ensure that Indian AI developers can access high-quality datasets without compromising citizens' data rights?
The DEPA mechanism is primarily about creating a secure framework through which data can be shared for research, artificial intelligence and other purposes without compromising privacy. We have already created AIKosh, which is operational and provides access to datasets for AI development. However, some datasets contain personally identifiable information, and even when such information is anonymised, there remains a possibility that anonymisation could be reversed.
This raises an important question: why should an entire dataset be supplied to a researcher when the requirement is only to undertake a particular analysis? We are therefore looking at a framework in which researchers can access and analyse data within a secure environment for a specified purpose. They can conduct their research, obtain their findings and take those findings away, but the underlying data remains within that protected environment.
The principle is similar to the reference sections of traditional libraries. There were books that could not be borrowed. Researchers could consult them, make notes and leave, but the material itself remained within the library. We are applying a comparable principle to digital datasets.
The National Health Authority is also adopting a similar approach for medical data, where protecting personal information is particularly important. Our objective is to establish good practices that can be adopted more widely by government agencies, private companies and research institutions. The framework could also be adopted internationally. It provides a way to facilitate data sharing and AI research while preserving privacy, anonymity and the conditions under which the data is made available. This is essential for enabling AI development without compromising citizens' rights.
India continues to depend substantially on imported GPUs and foreign technology providers for advanced AI computing. What is MeitY's long-term strategy for reducing these dependencies? Is India considering indigenous AI accelerators, domestic computing architectures and dedicated sovereign AI infrastructure for sensitive government and strategic applications?
The infrastructure required for artificial intelligence cannot be created overnight. We are taking a studied and comprehensive approach because the AI technology stack consists of five distinct layers, each of which needs to be addressed. The first layer is energy and power infrastructure. India is relatively well placed in this area, which is one reason why several international companies are interested in establishing data centres and AI computing facilities in the country.
The second layer is the chip itself, particularly the accelerator chips used for AI computing. Today, these chips are produced by a relatively small number of companies globally, and that concentration contributes to their high cost. Under India Semiconductor Mission 2.0, we intend to target the development of an indigenous GPU chip. The minister has also indicated this priority. We should have a GPU chip of our own. This will not happen overnight. It is a project that will require approximately 24 months, and we intend to work towards that objective.
The third layer is computing infrastructure, including servers and associated hardware. India already has companies capable of manufacturing servers. Companies such as VVDN, Kaynes and Avalon, among others, have developed manufacturing capabilities in this area. Several server components are also being manufactured domestically, supported by the Electronics Component Manufacturing Scheme and other initiatives.
There are still some components, particularly semiconductor chips, that are not manufactured entirely in India. Addressing those gaps is part of our strategy. Nevertheless, the capability to assemble and manufacture Indian servers already exists. C-DAC has developed its Rudra series of servers, including the AI-capable Rudra 3 series. These systems are increasingly being deployed, and technology transfer to Indian companies is helping expand domestic server manufacturing capabilities. We are also encouraging international companies establishing computing infrastructure in India to develop manufacturing facilities here.
The fourth layer consists of AI models. Under the IndiaAI Mission, we have supported more than 20 indigenous models. We hope that at least two of these will develop into frontier-level models, and we are continuing to support their development. At the same time, we recognise that large general-purpose models are not the only requirement. Smaller, specialised models can be highly effective in sectors such as agriculture, manufacturing, healthcare and education. These models can perform specific tasks efficiently while consuming less energy and computing power.
The fifth layer, and perhaps the most important from India's perspective, is applications. The real benefits of AI will emerge when it is deployed across businesses, industries and public services. Enterprise-level AI adoption remains limited. Individuals may be using AI tools, but many organisations have not integrated AI across their operations. Concerns about security, confidentiality and the possibility of sensitive business data being shared externally are among the reasons.
India has a significant opportunity to address these challenges and support enterprise-level AI adoption, not only domestically but internationally. By developing capabilities across all five layers, we can strengthen our technological resilience and eventually achieve greater strategic autonomy.
The government has approved Semicon 2.0 with an outlay of ₹1,27,500 crore, and notifications covering its six pillars were issued in August. How will this programme move India beyond semiconductor assembly and packaging towards advanced fabrication, indigenous chip design, equipment manufacturing and materials production? What measurable manufacturing milestones should India achieve over the next three to five years?
What we are attempting through India Semiconductor Mission 2.0 is to build the entire semiconductor ecosystem in the country. The first phase of the mission focused primarily on two areas: semiconductor wafer fabrication and the assembly and packaging of chips. We now recognise that India must develop capabilities across the entire value chain, beginning with chip design and extending to manufacturing equipment, materials, chemicals, advanced packaging and research and development.
Chip design is an important part of this strategy. Under the first phase, we introduced the Design Linked Incentive scheme, which has produced encouraging results. We identified more than 100 companies, of which 24 received assistance under the scheme. Subsequently, 15 of these 24 companies also attracted venture capital or other private investment. This demonstrates that government support can help companies develop technologies that subsequently attract commercial investment.
Under ISM 2.0, we intend to place greater emphasis on chip design, including strategically important chips such as AI accelerators. We are planning a higher allocation for this segment. We are also looking at different categories of chip design, including strategic chips and commercial chips that require support. The approach will include investment-based assistance rather than relying exclusively on grants. Support for electronic design automation tools will also continue.
The second major priority is developing the ecosystem of materials, machinery and equipment required for semiconductor manufacturing. A fabrication facility needs an extensive network of suppliers providing specialised gases, chemicals, materials and equipment. These supporting industries are essential for establishing a sustainable semiconductor manufacturing base. Under ISM 2.0, we intend to support their development. Several companies have already expressed interest and announced investment plans in these areas.
The third priority is to differentiate the support provided to various manufacturing activities. Semiconductor fabrication facilities will continue to receive substantial assistance, with support of 40%. Advanced packaging facilities will be eligible for assistance of 35%. The objective of providing higher support for advanced packaging is to encourage the industry to move towards more sophisticated technologies. Under the first phase, we attracted approximately eight or nine companies in conventional packaging. We now want the industry to progress to the next level.
Research and development will be another important component of the programme. We have established capabilities at the 28-nanometre technology node, but progressing towards more advanced nodes involves considerably higher costs. We therefore need to examine how research and technological development can help India move beyond 28 nanometres. This is an important area for future investment.
There are also emerging technologies, including display technologies and silicon photonics, where the government has supported research in various institutions. Some of these initiatives have produced promising results, and we now want to take them further towards practical and commercial outcomes. The overall objective of ISM 2.0 is to ensure that India develops a more complete and sustainable semiconductor ecosystem rather than remaining dependent on a few isolated segments of the manufacturing process.
Recent initiatives under the Chips to Startup programme and the Digital India RISC-V ecosystem have demonstrated growing domestic chip-design capabilities. What is the roadmap for commercialising indigenous processors such as SHAKTI and VEGA? Can India establish a significant market for domestically designed chips in government systems, defence, telecommunications, automotive electronics and industrial applications?
Both SHAKTI and VEGA processors have been developed indigenously using the RISC-V architecture. These developments demonstrate India's growing capabilities in processor design. As I mentioned, 24 companies received assistance under the Design Linked Incentive programme, and 15 of them subsequently raised venture capital funding. These companies are developing a range of products, including microcontrollers, flight controllers for drones, chips for smart meters and chips for CCTV cameras.
These are practical applications for which there is substantial demand. I believe the prospects for commercialising many of these products and taking them to a significantly higher level are strong. The government will continue to support this process. India also has a very large semiconductor market. Its current size is a little over $50 billion, and it is expected to approach $100 billion by 2030. The combined production from the semiconductor projects currently being developed will represent only a fraction of India's total consumption.
However, semiconductor chips serve different purposes, and there cannot be a single approach to developing the market. Different applications require different types of chips, with varying performance, reliability and security characteristics. We need to ensure that Indian chip-design companies have access to an adequate domestic market. That does not necessarily mean imposing restrictions across the entire consumer electronics market, because doing so could create unnecessary burdens.
There are, however, specific areas where the use of trusted and verified chips is particularly important from a security perspective. For example, CCTV cameras, industrial sensors, Internet of Things devices, drone flight controllers and smart meters can involve sensitive information. In such applications, data leakage or a compromise of system integrity can create serious security concerns.
It is therefore reasonable to ensure that the chips used in these systems come from trusted and verified sources. Indian companies could be encouraged to supply chips for such applications, provided their products meet the required standards of quality and performance. This approach can help establish a market for India-designed and India-made chips while also addressing important security requirements. It creates an opportunity for domestic semiconductor companies to develop commercially viable products without imposing unnecessary restrictions on the broader market.
The Digital Personal Data Protection framework is becoming increasingly important for government departments, digital platforms and private enterprises. How is MeitY preparing these organisations for compliance? What are the major challenges in implementing consent management, grievance redressal and regulatory requirements, particularly for smaller businesses and government departments?
We have provided an implementation period of 18 months, and the framework is intended to become operational by May 2027. The government had initially considered bringing forward the implementation date. However, several businesses represented that they would require more time to prepare for compliance. Taking these concerns into account, we have retained the implementation timeline.
We have stepped up our awareness campaign around the Digital Personal Data Protection Act, its provisions and the measures that organisations must take to comply with it. A number of awareness programmes are being conducted in Delhi and other centres across the country. We are also using social media, other communication channels and institutional mechanisms to ensure that businesses, government departments and other stakeholders understand their obligations.
This effort has already helped increase awareness of the legislation and the changes that organisations will need to introduce. The National Informatics Centre has also developed a product to assist government departments and other organisations handling substantial amounts of personal data in complying with the provisions of the Act. This tool is being widely publicised among government ministries.
The Cabinet Secretary has written to all secretaries in the Union government and the chief secretaries of states, emphasising the need to prioritise compliance with the DPDP Act. MeitY is supporting ministries and other government organisations in preparing for implementation. The objective is to ensure that institutions handling personal data are ready to meet their obligations when the relevant provisions become applicable.
At MeitY's June consultation on AI transformation of the IT services industry, you emphasised reskilling and technological upgrading rather than focusing exclusively on potential job losses. With generative and agentic AI transforming software development and business services, how should India's IT industry adapt its business model? What government-industry initiatives are being developed to protect competitiveness, create higher-value employment and prepare the workforce for AI-driven disruption?
NITI Aayog produced a paper last year examining the implications of AI for employment in India's IT and IT-enabled services sector. The paper pointed out that the sector currently employs approximately six million people. It anticipated that around two million existing jobs could become redundant because of technological changes, but that approximately four million additional jobs could be created. This would result in a net increase in employment.
The important question is how these new employment opportunities will emerge. I believe the principal opportunity lies in the deployment of AI across enterprises and organisations. This is an area where India has a substantial advantage. There is a parallel with the Y2K period, which created significant opportunities for the Indian IT industry. At that time, Indian companies were able to provide technology services to businesses around the world, and this contributed substantially to the growth of the sector.
AI could create a similar opportunity, although the nature of the work will be different. It is relatively easy to say that enterprises should adopt artificial intelligence. The actual process is much more complex. AI solutions must be designed around the specific requirements of each organisation. They must protect sensitive data, integrate with existing systems and be deployed effectively.
These are precisely the areas where India's IT industry can make an important contribution. We have a large pool of trained IT professionals with experience in developing, integrating and managing complex technology systems for organisations across the world. These capabilities can be applied to enterprise AI adoption on a substantial scale.
Indian companies can help businesses move from experimenting with AI tools to deploying AI throughout their operations. This will require technological adaptation and new skills, but it also presents a significant commercial opportunity. Our trained IT workforce gives India the capacity to undertake this work at scale and competitively. I believe this is where a major opportunity lies for the next phase of growth of India's IT services industry.
India plans indigenous GPU chips in 24 months, says MeitY Secretary S Krishnan on AI strategy
