India's ambitious push into semiconductor manufacturing requires a strategic pivot beyond conventional chip packaging, according to former Minister of State for Electronics & Information Technology, Rajeev Chandrasekhar. He argues that the nation must focus on higher-value activities like indigenous chip design, advanced fabrication, and dedicated research to truly capitalize on the evolving global semiconductor landscape, which is being reshaped by artificial intelligence.
Shifting Focus from Conventional Chip Assembly
Chandrasekhar highlighted significant progress in India's semiconductor program, citing 12 approved manufacturing units with a committed investment of ₹1.64 lakh crore, and five facilities already in production. The subsequent India Semiconductor Mission 2.0 has seen an additional ₹1.275 lakh crore. However, he cautioned that the nature of these investments is crucial. Nine of the twelve approved units are conventional ATMP/OSAT (assembly, testing, marking, and packaging) facilities, which represent the lower-margin end of the semiconductor value chain and face increasing competition.
While India has seen success in electronics manufacturing, partly due to the 'China+1' strategy leading to supply chain diversification (e.g., assembling 25-28% of iPhones globally), Chandrasekhar believes semiconductors are undergoing a different transformation. He noted, "Semiconductors are not experiencing supply-chain diversification. They are experiencing an architectural revolution."
AI's Architectural Revolution and New Opportunities
The rise of artificial intelligence is fundamentally altering the semiconductor industry, creating new demands for computing power. Chandrasekhar pointed to the sharp increase in Nvidia's data-center revenue as a clear indicator of the immense AI chip opportunity. This shift underscores the growing importance of semiconductor design and intellectual property in determining the value captured by companies.
India's Potential in AI Inference Chips
Chandrasekhar identified AI inference—the process of running trained AI models—as a significant area where India could establish a strong global position. While AI training often consolidates around ecosystems like Nvidia's CUDA and custom chips from hyperscalers, inference requirements are more diverse, spanning cloud, edge devices, defense, agriculture, and industrial applications. He asserted, "Inference is wide open." India possesses a substantial talent pool of 1.25 lakh chip-design engineers and can leverage open-source RISC-V technology through the DIR-V program to develop specialized chips without reliance on proprietary ARM licensing.
Proposals for India's Advanced Chip Future
To realize these higher-value ambitions, Chandrasekhar proposed several key initiatives:
- National Semiconductor Research Institute: A dedicated institute focused on process technology, design IP, and talent development.
- Chip Design Commercialisation Fund: A ₹1,000-crore fund, modeled on the National Investment and Infrastructure Fund, to support Indian fabless companies in financing development through commercial tape-out.
- Sovereign AI Inference-Chip Programmes: At least two such programs with guaranteed government offtake to stimulate domestic demand and capability.
Chandrasekhar concluded by emphasizing that India's next phase in semiconductors must prioritize building profound technological capability rather than merely expanding the number of approved facilities or focusing solely on headline investment figures. The goal, he stressed, is to ensure India does not miss this crucial industrial supercycle in the AI age.