HEMT device design

Fewer than a dozen nations can design a High-Electron-Mobility Transistor and turn it into hardware. India imports more than 90% of its semiconductor requirements as of 2025 — yet it is one of the eight countries with demonstrated HEMT capability.

A HEMT is a transistor built not from a single doped material but from a junction between two — most usefully an AlGaN/GaN heterostructure, where a sheet of exceptionally mobile electrons forms at the interface. That mobility is what lets these devices switch fast and handle high power at the frequencies radar and communications systems need. The difficulty lies everywhere at once: growing the crystal layers cleanly, engineering the gate so the device behaves predictably, managing the heat, and doing it reliably enough to qualify for military use. The physics is unforgiving of small defects.

India has cleared the hardest conceptual hurdle. The Indian Institute of Science, Bangalore, built the country's first enhancement-mode GaN HEMT, rated at 600V and 4A, using a novel aluminium-titanium oxide gate that addresses the stability and reliability problems of standard industrial techniques — a design published in IEEE Transactions on Electron Devices in June 2019. It sits at prototype stage, TRL 5.

The defence side has gone further into hardware. DRDO's Solid State Physics Laboratory, working under the BALRAM project, developed indigenous processes for 4-inch silicon-carbide wafers and GaN HEMTs up to 150W, with MMICs (integrated circuits combining several functions) up to 40W at X-band. That technology was transferred to GAETEC in Hyderabad, which established indigenous GaN-on-SiC MMIC production with limited capability for defence, aerospace and space applications — announced in November 2024. IIT Bombay continues underlying device-design research with DRDO and ISRO support.

The gap, then, is not design. It is volume. India has prototypes with competitive specifications and a limited-production line; it does not have high-volume output. The word attached to GAETEC's capability remains "limited". The structural reason lies in the layers beneath the transistor: SiC substrate manufacturing is still emerging, and reliability testing and qualification to military standards is emerging rather than mature. A device that works in the lab is not yet a device a radar programme can order by the thousand.

Closing that gap, the record suggests, requires sustained investment in epitaxial growth infrastructure, advanced lithography, thermal management, and radiation-hardening for space use. The proof of principle exists. The industrial base to repeat it at scale is what is still being built.