ASML begins operations in India and plans to hire young graduate engineers

Sandip Kumar, CEO of L&T Semiconductor Technologies, emphasized that forming a globally competitive semiconductor industry in India will require the creation of hundreds of domestic companies.

Speaking at the Semicon 2026 India conference, Kumar noted that due to the enormous scale of the global semiconductor market, it is impossible for one company to cover all product categories. He specified that there are over 20,000 semiconductor products, and no single company can produce even 2,000 of them, thus requiring the emergence of hundreds of enterprises.

Kumar believes that the appearance of more local chip manufacturers should be viewed as a necessary stage in building an ecosystem, rather than merely as competition. He added that part of L&T Semiconductor's efforts is directed towards fostering the development of a broader semiconductor ecosystem in India.

According to Kumar, India already possesses strong capabilities in digital and computing technologies, as well as analog and mixed-signal design, while its competencies in radio frequency technologies are developed to an acceptable level. However, significant gaps remain in the high-power semiconductors, memory, and optics segments, which he described as 'very weak' in terms of India's design and architecture capabilities.

Kumar also identified system architecture as another major deficiency. Unlike chip architecture, system architecture begins with understanding the needs of customers, markets, and the tasks the product must solve. He stated that 'the first macro gap, as far as I know, is system architecture, not chip architecture.'

Kumar explained that much of this knowledge comes through direct contact with customers and markets, something Indian semiconductor engineers have historically had fewer opportunities for because many large end-users are based outside the country. Furthermore, India lacks experience in transitioning from chip design to manufacturing, semiconductor supply chain management, and direct engagement with fabrication plants.

He noted that competition in the semiconductor industry is increasingly determined by the system-level architecture. Kumar stressed that the current boom in the sector is driven precisely by architecture, distinguishing it from previous periods. As individual chips reach physical production limits, the importance of technologies such as chiplets and advanced packaging increases. Demand is also stimulated by AI workloads, which require faster interconnects, co-packaged optics, and more efficient power delivery.

Kumar focused on the economics of the semiconductor business, where high gross margins necessitate significant reinvestment. He reported that gross margins of 50, 60, or 70% can be achieved in this field, but a substantial portion of these revenues must be directed toward developing next-generation products, as chip manufacturers are forced to constantly innovate amidst the replacement of existing products with new technologies.

L&T Semiconductor Technologies itself is expanding in the areas highlighted by Kumar. This company, which does not own its own fabs, has designed and started shipping over 40 products to 15–20 clients in the last year and a half, covering power modules, communications, power conversion, and computing. The company targets sectors such as energy, industrial applications, mobility, data centers, and AI.

The company is scaling its capabilities through acquisitions and partnerships. L&T Semiconductor acquired power module design assets from Fujitsu General Electronics, including intellectual property, design patents, and packaging expertise. It collaborates with Hon Young Semiconductor on silicon carbide wafer development and with Andes Technology on RISC-V processor platforms. A recent partnership was established with Taiwan's Azuremoto Technologies regarding silicon carbide power devices for AI data centers, as well as MOSFET rectifiers and power products. A multi-year agreement was also signed with Synopsys to strengthen power electronics design and modeling capabilities.

Kumar views India's ambitions in the semiconductor sphere as a long-term project that will take about 20 years. While the ecosystem is beginning to form, the next phase will involve deepening the technological base and creating products capable of competing globally. In Kumar's view, differentiation does not just mean being different, but that the product is superior compared to existing alternatives in global markets.