Amid a global reset in semiconductor strategy, India’s clean-slate advantage lies in frugal innovation, chiplet architecture, and agile microfabs. With no legacy constraints, we have
the freedom to shape the future rather than imitate the past.
“There is an Indian in every chip,” I often say, thinking of companies like Cadence, Synopsys, Applied Materials, and TSMC. Having been in this industry since my days at Texas Instruments (TI), I can say semiconductors are not new to India. We have contributed across the ecosystem—intellectual property (IP), electronic design automation (EDA) tools, and even equipment design, everything but manufacturing.
We may not yet have a homegrown semiconductor company, but this is our moment to build that ecosystem. The industry is set to double over the next 8-10 years, and India has a unique opportunity to shape its entry, not by replicating old models, but by innovating from first principles.
The ‘newness’ of India
Being a new player in semiconductors is actually an advantage; we do not carry the weight of legacy systems. I saw this first-hand at TI, where manufacturing and product development followed a more traditional, structured path.
Today, however, we have access to modern tools and approaches, and we no longer need to adhere to outdated methods.
India is known for its services, and I also ran a services company. Eventually, I recognised that controlling our future meant focusing on products, not services. Take SoCs (system-on-chip), for example: companies like Qualcomm attempted monolithic integration—putting everything into one giant chip. But that drove up cost, reduced yield, and customers disliked being locked into a single vendor.
Now the industry is shifting towards chiplets, which are modular components optimised for different processes and integrated at the package level. This approach allows greater flexibility, improved yields, and fosters innovation across various system elements.
India, unencumbered by legacy baggage, is well positioned to adopt this model.
Frugal engineering and teamwork
We are naturally frugal, raised to make every rupee count. With today’s exchange rates, our money stretches even further. But frugal engineering is not just about saving, it is about eliminating waste from the outset. That means designing products with efficient use of materials, resilient supply chains, and minimal resource loss. It is a mindset we grow up with.
One of India’s biggest advantages is scalable human talent. Rather than everyone building separate teams, there is immense potential in pooling skilled resources across functions.
Imagine two companies—Company A and Company B—working in similar spaces. They may compete, but there is room for collaboration, particularly where one possesses expertise the other lacks. From design and marketing to prototyping, many skills are not available internally. If companies are willing to share even a few expert hours per week, they can reduce costs and amplify collective success.
At a macro level, two companies supporting each other grow stronger together. This is especially important for Indian startups, where building a chip requires diverse capabilities, from design and testing to packaging and business strategy. Unlike software, where most aspects are handled internally, hardware demands collaboration across firms and functions.
However, one of India’s core challenges is a reluctance to collaborate. We are often raised to prioritise individual success. If someone else excels, the question becomes, “Why not you?” instead of “How can we succeed together?”
Yet chip development is not a one-person job. It requires multidisciplinary teamwork and the ability to recognise one another’s strengths. In more collaborative cultures, children are taught early to work in teams and celebrate others’ successes. That mindset, recognising that combining strengths yields better outcomes is essential for semiconductor innovation.
Frugal engineering also involves sharing resources. In manufacturing, we have been conditioned to believe that only mega fabs costing $10 billion or more can succeed. These vast facilities, established by dominant players, create barriers for frugal innovators.
But it was not always this way. At TI, we had our own fab, built not to fill capacity but to support the product. The product drove the fab, not the other way around. That logic has since reversed. We must reconsider our approach if we are to foster innovation with limited resources.
The strategy of fabs
Fabs determine which products can be run and how customers interact with you. From a fab’s perspective, the ideal scenario is continuous manufacturing of a single product. This minimises the need for innovation and process change. The preference is to follow a fixed process design kit (PDK), ensuring consistency and profitability.
Fabs seek large clients who can utilise full capacity, rather than smaller startups requesting limited wafer runs. They aim to maximise profit, often through high pricing, instead of pursuing lower-cost processes that depend on higher volume for similar revenue.
Globally, a few major players, TSMC, Nvidia, Intel (under restructuring), AMD, Micron, Samsung, among others, dominate the semiconductor industry. Fewer than ten companies effectively control the global market. Their scale and operating structures shape industry behaviour. While there is nothing inherently wrong with this model, it is centralised and resource-intensive.
India need not emulate this structure. We have an opportunity to disrupt it, guided by frugal engineering. This is not merely about spending less; it is about rethinking how we design, manufacture, and distribute semiconductors.
Rather than investing $10 billion into a mega fab in one city, we could build smaller, decentralised fabs across the country. A fab consumes vast resources and is highly energy-intensive. Built smaller, these fabs could be locally managed, more inclusive, and tailored to regional needs.
Historically, fabs have focused on logic and memory, optimised for maximum transistor density, as seen with companies like Nvidia. However, there is a growing shift towards subsystems: sensors, MEMS, analogue components, and chiplets. These form the ‘beyond compute’ ecosystem, complementing rather than replacing the processor. They do not require mega fabs. Distributed manufacturing is more appropriate.
Earlier, semiconductor development focused heavily on transistor modelling, layout, and scaling. Today, engineers must design for application-specific systems in real-world contexts, particularly in sensing and embedded domains. Understanding the end use is vital. This evolution gives India a timely opportunity.
With a large base of capable engineers and a range of local challenges, India is a natural testbed for innovation. My sister, a doctor trained in India, often says her time here made her a better physician due to the diversity and volume of cases she encountered. Similarly, Indian engineers face issues the world is trying to solve: food security, clean water, climate change. Indian engineering can deliver global solutions rooted in affordability and relevance.
At traditional firms like TI and National Semiconductor, product development typically began with a business case: What is the ROI, and can customers afford it? The focus was largely on affluent markets. India can adopt a different approach, which includes solving meaningful problems first, and then building affordable technology around them.
Consider a problem approached by Nvidia compared to an Indian startup. Nvidia might develop a high-end chip requiring advanced cooling systems. An Indian startup might offer an 80-90% effective solution at a fraction of the cost—practical and accessible.
Much of the semiconductor industry has been driven by the economics of silicon focussed on shrinking size and reducing cost. However, material science is opening new avenues. Smaller wafer sizes, such as 15.24cm (6-inch) or 20.32cm (8-inch), can still yield excellent outcomes when combined with new materials, especially in power electronics and high-voltage applications.
Opportunities in semiconductors are vast and not only in design but also in device physics, materials science, and system integration. For students in electronics or semiconductor fields, this is a remarkable time. Flexibility and curiosity will be key.
To conclude, I propose a shift towards agile, distributed fabs requiring less than $10 million in investment as a viable alternative to billion-dollar mega-projects. This model aligns with India’s needs and strengths, enabling us to lead with purpose-driven, affordable innovation for the world.
Startup ecosystem and government support
There are two government schemes: the Design Linked Incentive (DLI) and the Production Linked Incentive (PLI). DLI focuses on semiconductor design, while PLI supports component manufacturing, such as cables, displays, and PCBs. Together, they aim to build a complete electronics ecosystem. Innovation will be key not only in chip design but also in packaging, assembly, and system integration.
However, funding remains a challenge for Indian semiconductor startups. Last year, the government focused on manufacturing, including fabs and packaging. This year, the focus is shifting towards supporting startups through schemes promoted by MeitY. The DLI has increased funding and is also encouraging private investment.
The stated target is to fund 100 startups and scale at least five to $100 million in revenue within five years. The long-term goal is for two Indian companies to enter the global top 10 in a decade. Achieving this will require a complete ecosystem extending beyond just design talent.
Many Indian startups are design-focused, but they struggle to convert their ideas into final products. They often become IP companies due to the lack of manufacturing support. In contrast, Taiwan has produced numerous chip companies with significant revenues, including MediaTek. Their success stems from experience across the full value chain, something India must now develop.
Crossing the ‘valley of death’ from design to full product development requires infrastructure, funding, IP support, and partnerships. Collaborating with experienced Taiwanese companies could help accelerate this journey.
Approach to manufacturing
India must take incremental steps, starting with OSAT (outsourced semiconductor assembly and test), and testing is a pragmatic approach. These require lower investment and are comparatively less complex. Full-scale fabs can follow. There is no immediate need to pursue advanced nodes such as 3nm or 4nm. Mature nodes like 130nm and 180nm are more suitable starting points. These technologies are easier to implement and could succeed within a three- to five-year timeline.
The idea of microfabs, requiring investments of under $10 million, across multiple locations could raise awareness and encourage hands-on experience. These fabs can produce non-logic components, such as sensors or power devices, which are making them ideal for learning and experimentation.
India’s education system must also shift to enable more practical learning. Engineering students should be encouraged to tinker, build, test, and debug real circuits. This shift in mindset, alongside robust manufacturing, is essential.
Building for India and the world
Demand should not be a concern. While the market is global, success begins at home. Products must first solve local problems, be reliable, and easy to maintain. Once proven locally, they can be relevant globally. India has the opportunity to create solutions that are not only affordable but genuinely needed worldwide.
India stands at a semiconductor inflexion point, reminiscent of the 1980s when transistors became commercially viable. Now is the time to create millions of innovative products. The government is enabling this journey. For young engineers, this is a rare and exciting opportunity to contribute to a global transformation rooted in India.
This article is based on a session titled ‘Opportunities for Entrepreneurs in the Indian Semiconductor Industry,’ delivered by Raja Manickam, Founder and CEO of iVP Semiconductor, at IEW 2025, held at the KTPO Expo Centre, Bengaluru. It has been transcribed and curated by Shubha Mitra, journalist at EFY.