By Hitesh Bhardwaj, GM – Semiconductors & Devices, Mitsubishi Electric India
In 2030, the most consequential computer in a factory will not be the one on a manager’s desk. It will be embedded in the robotic arm on the floor: sensing, deciding, and correcting in milliseconds, without waiting for instructions from a server room halfway across the building.
This is the shift that a trillion-dollar semiconductor industry is quietly engineering.
Global semiconductor revenue reached $793 billion in 2025, registering 21% growth year-on-year. According to the latest projection from Gartner, the revenue for 2026 will be USD 1.32 trillion, of which AI semiconductors will represent about 30%. By 2029, AI processing semiconductors are projected to generate $438 billion in revenue, representing a 752% increase from the current figure. These are not abstract market statistics. They are a signal that computing is moving, physically, structurally, out of centralised data centres and into the machines, buildings, vehicles and infrastructure that constitute the physical economy.
The distinction worth paying attention to is not between old electronics and new ones. It is between systems that are connected and systems that are intelligent. A connected machine can report its condition. An intelligent one can interpret that condition, anticipate failure, and act; without human intervention. The semiconductor is what makes that difference possible. Advanced chip architectures, specialised AI processors and edge computing capabilities are allowing computation to happen where it is needed: at the source of the data, in real time.
Edge computing will be one of the crucial factors in this transformation. It can help to process data closer to where it’s generated, which will reduce latency and bandwidth requirements, and give faster response times in applications where latency is important. More importantly, it enables the intelligence to be part of the operating architecture of a system instead of an external computing service.
The industrial sector illustrates how quickly this transition is already underway. The International Federation of Robotics recorded 542,000 industrial robot installations worldwide in 2024 — more than twice the number deployed a decade earlier. Asia accounted for 74% of new deployments. Behind these robots are increasingly sophisticated semiconductor systems that integrate sensing, computation and AI into a single operating architecture. Machine vision, predictive maintenance, adaptive manufacturing, capabilities that once required significant external computing infrastructure, are moving onto the factory floor itself.
India sits at a significant inflection point in this story. The country’s semiconductor market is projected to grow at 19% CAGR and reach approximately $90 billion by FY2030, according to NITI Aayog. But the opportunity is larger than the chip market alone. India’s expanding electronics manufacturing base, its electric mobility sector, its rapidly scaling data infrastructure and its industrial automation ambitions all depend on the intelligent systems that advanced semiconductors make possible. The question for India is not whether to participate in this transition, it is how quickly domestic capability can be built to shape it rather than simply adopt it.
There is, however, a tension at the centre of this expansion that deserves more attention than it typically receives. The IEA projects global data centre electricity consumption will more than double to around 945 TWh by 2030. India’s own electricity demand is expected to grow at 6.4% annually through the decade. An ecosystem of intelligent systems, each drawing power, each processing continuously, places considerable pressure on energy infrastructure. This makes power semiconductors, energy-efficient chip architectures and intelligent energy management not peripheral concerns but foundational ones. The $1 trillion chip era will only be sustainable if efficiency scales alongside capability.
The deeper significance of this moment is not the market valuation, impressive as it is. It is that computing is becoming a structural property of the physical world — embedded in how machines operate, how buildings respond, how vehicles navigate, how infrastructure manages itself. The semiconductor is no longer just a component inside a device. It is becoming the intelligence layer of the physical economy.
By 2030, the systems that define industrial competitiveness, urban efficiency and everyday experience will be those that think for themselves. The race to build them is already underway.