By CA surekha Ahuja
If you are looking for the next ₹5–20 crore business, do not start by asking how to manufacture a chip. Ask what the chip industry will be unable to operate without.
SEMICON India 2026 is being held in New Delhi from 17 to 19 September with the theme Silicon to Systems. Building the Ecosystem, covering manufacturing, equipment, materials, packaging, testing, supply chain, design and related technologies. Applied Materials has also announced plans for a US$5 billion investment in India over the next decade, including R&D and supply-chain expansion.
For Indian founders, the important opportunity may therefore sit around the chip, not inside the chip.
The question is: What will semiconductor companies have to buy repeatedly, qualify carefully and replace reluctantly?
That is where a business can find a defensible entry point.
Who should enter?
Not everyone should start a semiconductor business.
| Business profile | Logical starting point |
|---|---|
| Existing industrial business | Adapt an existing capability and enter the supply chain |
| Existing technical-services business | Upgrade into specialised semiconductor applications |
| Technology startup | Solve a narrow, high-value problem with limited CAPEX |
| New manufacturing startup | Validate customer and qualification before building capacity |
| Capital-rich promoter | Invest only after technology, customer and economics are validated |
The key insight
Existing businesses may have the strongest starting position.
A company that already has engineering capability, machinery, technical manpower, quality systems and working capital may need an upgrade rather than an entirely new business.
A startup should generally look for a specific problem before looking for a factory.
Capital should be the last part of the equation, not the first.
Do not chase the chip. Chase the bottleneck.
The semiconductor ecosystem requires much more than fabs.
Look for dependencies around:
precision engineering | equipment maintenance | testing and calibration | specialised materials | clean-room services | water and utilities | packaging | logistics | automation | technical services
But need does not equal opportunity. A commercially usable opportunity requires:
customer + specification + qualification + recurring demand + economics.
The wrong model is: CAPEX → factory → product → customer
The disciplined model is: Customer problem → specification → prototype → qualification → order → CAPEX → scale
Customer first. Machinery later.
The existing-business advantage
An existing business should ask one question: What do we already know how to make, test, maintain or manage that the semiconductor ecosystem will increasingly need?
For example, an existing precision-engineering company may explore semiconductor-grade components.
A testing company may develop specialised testing or calibration capability.
An automation company may adapt its systems to semiconductor manufacturing.
A water-treatment company may explore specialised process-water requirements.
The objective is not to abandon the existing business.
It is to move one step closer to a high-value supply chain.
That can mean lower incremental CAPEX, existing manpower, established systems and potentially faster commercial validation.
What about a startup?
A startup should not enter because the semiconductor market is large.
It should enter because it has found a specific problem worth paying to solve.
Potential areas include: inspection | testing | automation | monitoring | specialised software | equipment support | process technology | technical services
The startup sequence should be: Problem → prototype → pilot → customer validation → qualification → commercial order
Only then should substantial capital be committed.
What about a capital-intensive project?
This is where the greatest discipline is required.
Before committing substantial CAPEX, establish:
Who is the customer?
What exactly will they buy?
What is the specification?
How long is qualification?
What is the expected volume?
What is the payment cycle?
Can the facility serve another industry?
Then stress-test:
lower volume + delayed approval + higher CAPEX + slower collections + higher rejection + delayed incentive.
If the project fails under reasonable downside assumptions, the answer is not to increase the investment.
The Semiconductor Opportunity Test
Before investing, get six answers:
| Test | Question |
|---|---|
| Customer | Who will pay? |
| Specification | What exactly must be supplied? |
| Qualification | What must be proven before approval? |
| Demand | Is there repeat demand? |
| Cash | Can the business survive the qualification and ramp-up period? |
| Second market | Can the capability serve another industry? |
If these six answers are not clear, the business is not ready for major CAPEX.
The CA's role is before the investment
The financial review should cover:
GST — classification, ITC, exports and job work.
Customs — classification, valuation, duties and exemptions.
Income tax — depreciation, R&D expenditure, capitalisation and applicable incentives.
FEMA and transfer pricing — where foreign investment, technology, royalties or associated enterprises are involved.
And one discipline is critical:
Prepare the project both with and without incentives.
An incentive can improve a viable project.
It should not be used to disguise an unviable one.
Get qualified before everyone starts looking
An existing business considering this opportunity should create a Semiconductor Vendor Readiness File: capability | machinery | technical team | quality systems | testing | capacity | compliance | traceability | financial strength | safety | business continuity
The objective is simple: Do not wait for the procurement opportunity to appear before becoming vendor-ready.
Qualification can itself become the competitive barrier.
The five mistakes to avoid
1. Confusing an investment announcement with an order.
2. Using the total semiconductor market as your revenue projection.
3. Assuming the lowest price will win qualification.
4. Calculating profit without calculating working capital.
5. Building capacity before validating demand.
The most expensive mistake is usually the last one.
The 90-Day Founder Move
Days 1–15: Identify 10 supply-chain bottlenecks.
Days 16–30: Select three based on customer pain, qualification barrier, recurring demand and capital requirement.
Days 31–45: Meet actual customers and obtain specifications.
Days 46–60: Test technical and financial feasibility.
Days 61–75: Pilot and begin qualification.
Days 76–90: Make the investment decision.
Not before.
Final Words
The semiconductor opportunity should not be reduced to:
“India is building semiconductor capacity. Therefore, build a semiconductor factory.”
That is too simplistic.
For an existing business, the opportunity may be to upgrade an existing capability and become a qualified supplier.
For a startup, it may be to solve one specialised problem without carrying unnecessary CAPEX.
For a capital-intensive promoter, the sequence should be:
customer → technology → specification → qualification → economics → capital.
Not: capital → factory → hope.
The real competitive question is: Which dependency will become more important as India's semiconductor ecosystem expands, and can your business become difficult to replace?
Find the bottleneck before the crowd finds it.
Validate the customer before buying the machinery.
Build the capability before scaling the capital.
The next semiconductor opportunity may not be the company making the chip.
It may be the company that the chipmaker cannot afford to operate without.
