Archive · 5 Aug 2026
India's semiconductor push could define its AI ambitions
Artificial intelligence, or AI, is like a very smart computer that needs powerful machines to think.
Those machines depend on tiny parts called chips.
A new report from Crisil Intelligence says AI rests on five layers: electricity, semiconductors, infrastructure, foundation models, and applications.
Semiconductors, the chips, are one of the most important layers.
India plays a big role in assembling, packaging, testing, and designing chips, but it does not yet make them fully from scratch.
That is starting to change, as Micron Technology is building a big chip facility in Sanand, Gujarat, with more than $800 million.
A joint venture between Tata Electronics and Powerchip is also planning to make 28-nanometre chips in Dholera, Gujarat.
Making more chips in India can help the country become stronger in the global AI race.
The report reminds us that a chip company is not automatically an AI company.
It also says most AI investment, about 85%, goes to infrastructure and applications, while only 15% goes to foundation models.
A Crisil Intelligence report describes AI as a five-layer 'penta-stack' economy in which semiconductors occupy one of the most critical positions.
India's semiconductor ecosystem is expanding in assembly, testing, packaging, and design, but advanced fabrication is still developing.
Micron Technology is investing more than $800 million in a chip assembly and testing facility in Sanand, Gujarat.
The Tata Electronics-Powerchip Semiconductor Manufacturing Corporation joint venture at Dholera, Gujarat, aims to produce 28-nanometre chips.
AI infrastructure and applications attracted 85% of tracked AI investment capital, while foundation models attracted the remaining 15%.
- Who
- Crisil Intelligence published the report; it assesses Indian companies including Micron Technology, Tata Electronics, Powerchip Semiconductor Manufacturing Corporation, Adani Green Energy, Tata Power Solar, and ReNew Power.
- What
- AI is described as a five-layer ecosystem with semiconductors as a critical layer, and India is strengthening chip assembly, packaging, design, and energy capabilities but has not yet become a full semiconductor manufacturing hub.
- When
- Not specified in the article.
- Where
- India, specifically Sanand and Dholera in Gujarat.
- Why
- Advanced chips are the computing backbone that trains and runs AI models, so strengthening semiconductor capabilities is seen as key to India's role in the global AI race.
- Report publisher
- Crisil Intelligence
- AI ecosystem layers
- Electricity, semiconductors, infrastructure, foundation models, applications
- Semiconductor facilities
- Micron assembly and testing plant in Sanand, Gujarat; Tata Electronics-Powerchip fabrication project in Dholera, Gujarat
- Micron investment
- More than $800 million
- Chip node
- 28-nanometre
- AI investment split
- 85% infrastructure and applications; 15% foundation models
- Language models cited
- ChatGPT, Gemini, Llama, Claude, DeepSeek