Can India manufacture silicon carbide power semiconductors at scale?

Every silicon carbide power device in India is imported. As of 2025, the country buys 100% of its SiC devices from abroad, part of a broader dependence in which more than 90% of all semiconductor components come from overseas.

Silicon carbide is a wide-bandgap material that switches high power at high efficiency — the backbone of electric-vehicle powertrains, renewable-energy inverters, and industrial systems. SiC MOSFETs alone accounted for more than 41% of a global power-device market worth $2.99 billion in 2024, projected to reach $12.65 billion by 2032. Only five nations — the United States, Europe, Japan, South Korea and China — supply this market at scale.

The difficulty is physical, not administrative. Growing a single SiC crystal ingot demands roughly 2,600°C and pressures above 350 MPa; growth is slow, capacity limited, and quality unstable. The material's hardness — second only to diamond — makes slicing and polishing a yield bottleneck, and substrate defects must be held below 1,000 dislocations per square centimetre. Substrate costs account for about 45% of total device cost. This is why the capability sits with so few.

India stands at the foot of the climb, but it has started. DRDO's Solid State Physics Laboratory has demonstrated indigenous 4-inch SiC wafers and established processes for 4H-SiC bulk single-crystal growth and epi-ready wafer fabrication — a capability it describes as unique in India, though at laboratory scale. Continental Device India Limited became the first Indian firm to package SiC devices, opening an auto-grade line at Mohali in September 2023 using imported wafers, and won ISM approval in August 2025 to expand into high-power discrete production. Three commercial fabs are under construction: SiCSem in Odisha (Rs 20 billion, targeting 60,000 wafers a year by 2027–28), RIR Power Electronics in Bhubaneswar (groundbreaking September 2024), and Indichip's planned $1.4 billion fab in Kurnool.

The gap persists because scale requires the full stack, and the foundation is thin. Ingot growth and bulk-material synthesis — the critical first step — has no domestic base. Wafer slicing remains emerging. Downstream processes such as epitaxy, ion implantation and packaging have been demonstrated but not proven at volume. SiCSem's plant relies on technology and equipment from UK-based Clas-SiC.

What it would take is the vertical chain, from ingot to qualified device, operating together. India has demonstrated the individual steps and broken ground on the fabs. Converting demonstration into reliable, defect-controlled volume — meeting automotive qualification standards — is the mountain still ahead.