This post is also available in:
עברית (Hebrew)
Gallium nitride has become an important material for high-performance radar, communications and electronic warfare systems because it can handle far more radio-frequency power than older semiconductor technologies. But there is a catch: as engineers push more power through GaN transistors, they generate intense heat in extremely small areas. If that heat cannot escape quickly enough, performance and reliability suffer.
A DARPA research program is trying to remove that bottleneck by redesigning how heat moves through the electronics themselves. BAE Systems’ FAST Labs has completed the first phase of the Technologies for Heat Removal in Electronics at the Device Scale (THREADS) program and will continue into Phase 2, where researchers will further develop techniques for cooling high-power RF devices at the transistor level.
The approach differs from simply attaching a larger heatsink or cooling system. The program focuses on reducing thermal resistance inside the semiconductor device, which is the resistance that prevents heat from flowing away from the hottest regions. Researchers must accomplish this without interfering with the electrical properties that allow the transistor to generate and amplify high-frequency signals.
According to Interesting Engineering, that challenge is particularly important for GaN, a wide-bandgap semiconductor capable of operating at high frequencies and power densities. It is already used in technologies such as active electronically scanned array (AESA) radar, where large numbers of small transmit-and-receive modules electronically steer radar beams without mechanically moving the antenna.
According to the program, GaN has already enabled more than a fivefold increase in power density compared with previous transistor technologies. It is targeting an eightfold reduction in thermal resistance and power densities reaching 81 watts per millimeter in X-band transistor and amplifier test devices.
During Phase 1, program participants achieved approximately five times the RF power density of current state-of-the-art devices while maintaining operationally relevant reliability, according to the program’s latest update. The agency says that level of improvement corresponds theoretically to roughly twice the radar range.
That does not mean installing a program-developed transistor will automatically double the range of every radar. Detection distance also depends on antenna design, target characteristics, atmospheric conditions and signal processing. However, increasing usable RF power gives engineers considerably more room to improve complete systems.
For defense applications, the benefits could extend beyond radar. Higher-power GaN electronics could support stronger electronic warfare effects and more capable communications while avoiding the size and weight penalties associated with larger cooling equipment. Improved thermal management could also allow compact platforms such as aircraft and unmanned systems to carry RF systems that would otherwise generate too much heat.
The program builds on earlier work exploring approaches such as GaN-on-diamond transistors, which use diamond’s high thermal conductivity to draw heat away from semiconductor junctions.
Phase 2 will push the technology further toward practical devices. The objective is ultimately straightforward: let GaN electronics use more of the power they are physically capable of producing without heat becoming the limiting factor.
