For decades, extending radar range meant finding ways to see farther: more power, better antennas, smarter signal processing. Now the challenge is increasingly what happens after all that power goes in. The radio-frequency components at the heart of modern radars are becoming smaller and more powerful, but they also generate more heat. Unless engineers can carry that heat away, some of the performance packed into the hardware remains out of reach.

A radar that can see an aircraft, missile or drone farther away can give operators precious extra seconds to detect, track and respond. More RF power can also push electronic warfare effects farther and strengthen communications links. The same underlying electronics drive all three capabilities, making the ability to manage heat an increasingly important part of the military technology race. DARPA has identified thermal management as a constraint on the performance of radar, electronic warfare and communications systems.

Military electronics have shifted to gallium nitride (GaN), which packs more RF power into a smaller device than earlier technology, a major advantage for active electronically scanned array (AESA) radars. But higher power density also concentrates more heat in a smaller area. A GaN transistor can theoretically output more power than the system can sustain before rising temperature degrades performance and shortens its lifespan. Today’s RF systems typically run below their transistors’ electronic limits for this reason. That’s the problem behind the Pentagon’s renewed push into cooling for RF electronics: removing heat close to where it’s generated, inside and around the transistor, before it can bottleneck the system.

DARPA’s leading effort is Technologies for Heat Removal in Electronics at the Device Scale, or THREADS, which aims to cut thermal resistance inside RF transistors and move heat away from the hottest parts of high-power devices. Its targets include an eightfold reduction in thermal resistance and a power density of 81 watts per millimetre for X-band transistors and power amplifiers . In June, DARPA reported that Phase I performers achieved roughly a fivefold increase in RF power density over today’s state-of-the-art devices, a gain the agency said translates into about double the radar range while preserving required device lifetimes. Phase II is focused on pushing power and thermal performance further, alongside a DARPA study into which RF power levels would be most useful for military platforms.

BAE Systems is among the contractors advancing to that phase. On August 12, the company announced that its FAST Labs organisation had completed Phase I and received continued support for Phase II, working at its Microelectronics Center in Nashua, New Hampshire, with Modern Microsystems and researchers from Penn State, Stanford, Notre Dame and the University of Texas at Dallas. BAE says the resulting thermal improvements could eventually enable nearly three times the range of RF systems.

These figures indicate what improved power density can unlock rather than offering a direct formula for radar range, which also depends on antenna aperture, frequency, receiver sensitivity, waveform design, signal processing and the target itself. Extra RF power mainly gives engineers more signal strength and design margin to work with.The heat originates in a tiny region near the active junction of a GaN transistor, where electrical power converts into RF energy. From there it must pass through the semiconductor and several material layers before reaching the cooling system, and each interface adds resistance. At modest power levels this is manageable; at higher densities, those microscopic barriers can end up setting the operating temperature of the device. Much of the engineering response focuses on shortening that path: new heat-spreading structures, thermal interface materials and cooling architectures designed to move heat away from the active device more efficiently. Some approaches embed high-conductivity materials directly in the device package; others restructure the semiconductor to reduce resistance as heat moves outward. Liquid cooling goes further still, bringing coolant near the junction to shorten the distance heat must travel, though it introduces its own challenges around pumps, pressure, leakage, reliability and power draw.

Microchannel cooling, which routes fluid pathways into very small spaces, has drawn particular research interest for extracting heat from exactly where RF devices generate it most intensely. The work remains experimental, but it shows how the cooling problem has moved from the radar enclosure down into the architecture of the electronics themselves.

Military radars already pack substantial capability into tight spaces. An AESA can house many transmit-receive modules across its antenna face, and even small individual heat loads add up. Bulkier cooling hardware risks consuming the space, weight and power savings that advanced semiconductors were meant to deliver. The most valuable cooling technology may be those that cut the thermal burden without adding a large new cooling system: heat spreaders near the transistor, new junction materials, microscopic fluid channels, or some combination of these in a single package. DARPA is also developing thin-film heat spreaders for high-power-density electronics as part of this effort.

A radar cannot engage what it cannot detect, and earlier detection gives the rest of a defence network more time to build a track and coordinate a response. A radar that sustains greater RF output can push its detection boundary outward, which matters most against targets that are small, hard to detect, or operating in a contested electromagnetic environment. Cooling has effectively become a performance technology. Removing enough heat from the transistor lets engineers tap power the semiconductor was already electrically capable of producing but couldn’t sustain, the same premise driving THREADS: excess heat currently cuts into both the performance and lifetime of high-power RF systems .

This technology is still maturing, and the gap between a lab demonstration and a fielded radar remains wide. A new thermal architecture must be manufacturable at volume, hold up through temperature cycling and vibration, run reliably for years, and preserve its advantages once integrated into a complete RF module rather than tested in isolation.

The next meaningful advance in radar may therefore come from a change that is invisible to the operator. The antenna may look largely the same, while the real breakthrough takes place underneath it, where engineers are finding faster ways to move heat away from the transistor. If they can turn more of a GaN device’s electrical capacity into sustained RF output, they could give future radars more reach, greater operating margin and more time to respond without demanding a proportionally larger machine.