Semiconductor Chips Will Come, But Only After India Has Enough Electricity

The hoardings went up quickly. “Semiconductors Made in India”. The Prime Minister has used the phrase in Davos, bilateral summits and previous editions of SemiconIndia. The India Semiconductor Mission (ISM) was launched with an investment of ₹76,000 crore, and has since then approved projects worth over ₹1.60 lakh crore across six states.

The Dholera manufacturing facility of Tata Electronics. Micron has an ATMP factory in Sanand. HCL and Foxconn have joined hands in Uttar Pradesh. The momentum looks real on paper. In fact, the ground beneath these aims is less stable than the press releases imply – and electricity is the first thing to upset it.

Geometry, Ambitious

India’s semiconductor market is projected to grow at a compound annual rate of more than 12% to around $161 billion by 2033, from $53 billion in 2024. The government’s drive has its logic. The pandemic, the US-China tech war and lingering fears about the Taiwan Strait have broken global supply chains, leaving every major economy desperate to diversify manufacturing. With the large domestic demand base, growing skill pool in design and a government willing to provide 50% fiscal support on project costs, India has become a credible alternative destination for chip investment.

Dholera’s greenfield fabrication facility, set up in collaboration with Taiwan’s Powerchip Semiconductor Manufacturing Corporation (PSMC), has an investment of about ₹91,000 crore and is expected to manufacture 50,000 wafers a month.

Micron has begun commercial production at its assembly and testing plant in Sanand, marking India’s first major foray into large-scale memory packaging. ISM 2.0 has expanded its scope from packaging to equipment, design intellectual property, and research and development. All of the above is accurate. But every one of these facilities will face the same bottleneck: steady, reliable, industrial-grade electrical power.

What a Fab Really Needs

Semiconductor fabrication plants are not like typical factories. A single sophisticated fab requires 100-150 megawatts of electricity all the time. This is equivalent to running a mid-sized city of one lakh homes, 24x7, without a break. But scale is only part of the picture. What matters is not only the quantity of power, but the quality of power. Voltage fluctuations, micro-interruptions or even momentary disturbances can wipe out whole batches of wafers worth crores of dollars in a matter of seconds. That’s why fabs worldwide are designed to run on grids engineered to be nearly 100% reliable over decades.

The semiconductor industry is one of the most energy intensive manufacturing industries in the world. According to published data, the global semiconductor industry consumed 149 billion kilowatt-hours in 2021, enough to power a city of over 25 million people for a year. Since then, that number has only grown as chip layouts have become more complex and node sizes have shrunk to single-digit nanometers. The lower the node, the more process steps are needed and the more energy each step uses.

Water makes things harder. A typical semiconductor fab uses over five million gallons of ultra-pure water a day. To get that kind of purity, it takes over eight million gallons of city water a day. This is not a minor engineering footnote in a country that already dedicates 70% of its water withdrawals to agriculture and where per capita water supply is declining in places; it is a fundamental conflict that has barely made it into mainstream policy debate.

India’s Power Grid: Achievements and Hurdles

India has made real advances in power generation. The country has moved from being power-starved to power-surplus with an installed capacity of over 446 gigawatts. The development of renewable energy, in particular solar, has been substantial and internationally recognised. But overall capacity is not the problem for semiconductor production. It is the consistency, localised trustworthiness and contractual certainty.

In industrialised areas, blackouts still occur during heat waves. But load-shedding continues in some states, although less than at its height 10 years ago. The voltage swings that would be acceptable in a textile mill or a plastics plant are disastrous in a cleanroom making 28-nanometer chips.

In addition, grid stability varies significantly from state to state. Gujarat has invested heavily in renewable infrastructure and currently has the most approved semiconductor projects. But Gujarat’s grid, like any grid increasingly turning to intermittent solar generation without the right storage, has reliability risks that fabs can’t afford.

There is also the question of guarantees of policy. Semiconductor companies that spend billions of dollars on a plant need contracts that assure them of supply – the kind of contract that has legal consequences and is not just the whim of a minister. Taiwan, South Korea, and even US states vying for the attention of TSMC and Intel have made similar promises backed by decades of grid engineering. While talks are on, India has not institutionalised this on a large scale.

Renewable Energy is Not Every Answer

India’s renewable energy expansion is widely touted as a solution to both energy insecurity and carbon obligation. Solar and wind power can meet a fraction of a factory’s energy needs, and Intel’s Bengaluru campus already has on-site solar PV systems.

However, semiconductor fabs cannot operate on intermittent generation without an extensive battery storage infrastructure or hybrid grid connection for base load stability. Battery storage at the scale a fab needs is still expensive, and India’s storage buildout, although growing, is far behind the pace of solar installation.

Also, the carbon footprint of semiconductor manufacturing, which is 95.8% globally dominated by grid electricity and fossil fuels, cannot be reduced by routing one plant through a renewable energy certificate. The upstream supply chain has its own carbon and energy costs—chemical manufacture, equipment manufacture, logistics, etc. For India to be serious about putting its semiconductor ecosystem inside the developing ESG-compliance frameworks that global buyers and investors are rushing to demand, it would require a far more systemic energy change than what is currently being proposed.

What Needs To Be Done

None of this is to say that India shouldn’t develop its own semiconductor industry. Strategic self-sufficiency in chips is a legitimate national goal, and the current window of geopolitical opportunity—with global supply chains in disarray—is both real and fleeting. It is an argument for openness about the conditions that distinguish intention from achievement.

The government should see power infrastructure for the semiconductor cluster as a critical part of the Semicon India Programme, not an afterthought at the state level. This includes ring-fenced transmission capacity, dedicated grid corridors with built-in redundancy, contractual supply guarantees with enforceable fines, and investments in storage technologies to address the intermittent nature of renewable inputs. This requires engaging with state power boards, rather than assuming private actors will create captive generation and absorb the costs.

It also means a more honest accounting of public money and who pays the price when heavy industry takes too much from a strained grid. The story of semiconductors in India need not be told in triumphalist terms only. The chips will come, but only after the electricity comes first.

The writer is a columnist and climate researcher with experience in political research analysis. ESG research and energy policy. The views are personal.