Fab Dreams, Design Reality

From missed opportunities in the 1960s to today’s fab ambitions, India’s semiconductor quest hinges on aligning world-class chip design with homegrown manufacturing.

In the context of the ‘Make in India’ initiative, our ambition of becoming an independent semiconductor‑manufacturing hub has become a national discourse. Having spent the last decade and a half in hands‑on micro-manufacturing, both in academic and industrial R&D with a few patents in the field, I know how crucial this field is for a country’s scientific and technological progress, its job market and above all, its long‑term economic sovereignty.

The political history of Taiwan’s rise as a manufacturing giant, and the way global consumer‑electronics companies now depend almost existentially on TSMC, is no longer a mystery. This alone explains China’s all‑in geopolitical interest in Taiwan. The post‑COVID tug‑of‑war in the chip world has also made it clear that a monopoly in semiconductor manufacturing can easily turn a technocratic state into what Neil Postman once termed a global ‘Technopoly.’ 

So, our aspiration to become a manufacturing powerhouse is logical. But how far is India’s chip platform from the global benchmark? And what does our own technocratic history suggest?

Chip Supply Chain

For perspective, and bypassing the technical intricacies of chip fabrication, let's examine a familiar product - the iPhone - and understand its internal technological map and supply‑chain logic.

A modern smartphone rests on three core components: the memory board for computation and storage, the logic board with sensors and filters that manage data flow, and the radio-frequency board that connects the device to communication bands. Because smartphone makers operate in a B2C market, these components are typically sourced from a network of B2B suppliers.

Without Bosch’s accelerometer/gyroscope, the iPhone’s GPS wouldn’t function; without Cirrus Logic’s audio amplifier, its multimedia system would be mute. Elementary, as Mr. Homes would point.

Apple assembles these components - mostly in Asian facilities - into the final product. But the design of each chip remains Apple’s intellectual property. The B2B manufacturers are contractually bound from reusing or selling those designs. These design architectures are the lifeblood of deep‑tech corporate R&D.

This model is the ‘fabless model’ where the B2C company owns the design, while manufacturing is outsourced to specialized B2B entities. Maintaining fabs, cleanrooms, and process infrastructure requires enormous capital and human resources, so outsourcing is a rational choice.

The manufacturers themselves convert these designs into physical chips on wafer scale. Their process innovation determines yield, repeatability, performance, and ultimately sets the B2C cost of goods or the cost of smartphones sold. This fabrication-process R&D is slow and expensive; it takes years to turn a design into a market‑ready component. For perspective, the global semiconductor industry today is valued at roughly three trillion dollars or Rs. 38 lakh crore.

Semiconductor Ecosystem

India’s national aspiration, however nebulous it may sometimes appear, is to build a semiconductor ecosystem that includes genuine, independently operating pure‑play foundries. Policymakers and scientific committees have long argued that even a handful of functioning fabs could help India have a piece of that 3TD pie reduce our dependence on an Asia‑centric manufacturing landscape. But where, historically and technologically, do we actually stand?

India’s early semiconductor ambitions date to the 1960s, when Fairchild Semiconductor considered setting up a fabrication unit in the country. The plan fell through amid bureaucratic hurdles, and the firm instead moved to Southeast Asia. After the Sino‑Indian War, Bharat Electronics Limited established a silicon–germanium device facility whose products even found overseas buyers, showing India was not initially absent from the race.

But as Taiwan, South Korea and later China rapidly scaled semiconductor manufacturing with strong policy backing, India’s early efforts faltered. Subsequent ventures including Metkem Silicon Ltd. and VEL struggled with funding and technology gaps. The most promising initiative, Semiconductor Complex Limited, suffered a crippling blow when a fire in 1989 set India’s chip ambitions back by years, if not decades.

Cut to 2013–14: the Government of India approved two large semiconductor wafer‑fab proposals; a consortium that included HSMC and international partners such as IBM. The projected investments were tens of thousands of crores. Yet neither took off. Financing challenges, ecosystem gaps, and commercial viability concerns proved cumulatively detrimental.

No matter how passionately we advocate for India’s fab ambitions, building a modern fab today and producing chips suitable for an iPhone‑class product is not an immediate possibility. Even with full government support, achieving Southeast‑Asia‑level process excellence, and developing automation‑driven, repeatable manufacturing will take at least one to two decades. Moreover, the continuous improvement trajectory of Asian process integrations adds to the challenge.

Automation and AI‑enabled manufacturing is now the global norm because it minimizes performance variation between chips of the same design. Big‑tech manufacturers are aggressively moving toward operator independence. This raises a critical question: Do sweeping promises of job creation make sense? Highly trained scientists and engineers will certainly find opportunities, but grassroots-scale employment so advertised is highly unlikely. As Dwaipayan Banerjee notes in ‘Computing in the Age of Decolonization,’ the sector actually lost nearly a million workers between 2016–17 and 2022–23.

We all want our country to become self‑reliant in chip manufacturing. But until full process independence is achieved, what we can realistically do is assemble, not manufacture. And that is precisely what we have been doing.

Meanwhile, published data shows that 20 percent of chip architectures at Apple, Samsung, NXP, Qualcomm, Intel, and Texas Instruments etc. are designed by Indians. If this intellectual strength were consolidated on Indian soil with proper compensation, wouldn’t we already have our own homegrown giants?

Until we reach world‑class fab independence, India should focus on becoming the dominant force in chip design. We must not lose our existing global edge in sensor/circuit design and IP. If we can preserve our strength in chip design while steadily building manufacturing capacity, the moment when design and fabrication finally converge on Indian soil could mark the true maturation of its silicon revolution and the emergence of India as a formidable global tech power.

(The writer is a Lead Process Engineer with GE HealthCare in France and a columnist with four books to his credit. Views personal.)