Chinese researchers say they have cracked a long-standing thermal bottleneck in advanced radar electronics, claiming a leap in performance without bigger antennas, heavier cooling gear or extra power draw. If verified and scaled, the work could hand Beijing a tangible head start in so‑called “super‑radars”.
Heat, not stealth, has been capping radar performance
Modern military radars rarely fail because their signals are too weak. They fail because their electronics overheat first. Every time an active electronically scanned array (AESA) radar pushes more power through its transmit modules, its gallium nitride (GaN) chips run hotter. At some point, engineers must back off to keep the hardware alive.
GaN has become the backbone of cutting-edge systems because it tolerates higher voltages and frequencies than older gallium arsenide components. Chinese fighters like the J‑20 and J‑35 are already reported to field GaN-based AESA radars, and the US has been moving GaN modules into variants of the F‑35 and ground-based systems.
The same physics that delivers that punch also brings pain. In the X and Ka bands used for fire control, long-range tracking and satellite links, GaN devices pour out heat faster than traditional cooling structures can pull it away.
Engineers have spent twenty years hitting the same ceiling: not an electronic limit, but a thermal one baked into the chip’s internal layers.
Past design tweaks focused on transistor geometry or packaging. The Chinese team instead went after an obscure internal interface where heat was quietly getting stuck.
The “invisible layer” that was holding everything back
A bottleneck buried inside the chip
At the heart of the new work by Xidian University lies a thin bonding layer within the GaN radio-frequency power device. This layer connects different semiconductor materials but is buried too deep to see with the naked eye.
Traditionally, engineers use aluminium nitride (AlN) at this interface. It performs well electrically, yet its crystal growth tends to form messy microscopic islands. From an electrical standpoint that can be acceptable. From a thermal standpoint it is terrible.
Those disordered islands act like speed bumps for phonons, the quantum units that carry heat through a solid. As a device ages under high load, the interface becomes even more resistant to heat flow. The radar module must then run at lower power or risk failure.
➡️ France loses a €3.2 billion Rafale deal after a last-minute reversal
➡️ For every dessert, its apple! The ultimate guide to finding the right apple to use
➡️ Trump threatens to impose photo ID for voters for midterm elections | Donald Trump
The group led by researcher Zhou Hong claims to have forced this layer to grow as a smooth, uniform film instead of a sea of micro‑islands. In effect, they turned a patchy, high-resistance bridge into a straight thermal motorway leading out of the active region of the chip.
By cleaning up a layer only nanometres thick, the team reports cutting thermal resistance by about one third.
That single figure matters. Thermal resistance tells you how much the device’s temperature rises for each watt of power. When it falls, engineers can either increase output power for the same temperature, or hold power constant and slash cooling complexity.
What “40% more performance” really buys a radar
According to Xidian University, the improved interface delivers roughly 40% higher radar performance without changes in chip area or energy consumption. That does not mean a raw 40% jump in range, but it unlocks several key gains for system designers:
- greater detection range without enlarging the antenna
- finer target separation at long distances
- stronger resilience against jamming and clutter
- faster update rates against high-speed threats
For a stealth fighter, this can translate into “seeing first” while emitting less often or at lower power, which helps keep the aircraft harder to detect. For ground-based air defence radars, it means covering wider volumes of airspace with the same hardware footprint.
China’s researchers argue that the gain comes from better thermal plumbing, not brute-force power, which keeps size and weight in check for aircraft integration.
On mobile platforms – from drones to naval vessels – that matters. Space and power budgets are tight. More capable radar without thicker cooling pipes or bigger generators is a direct operational advantage.
China’s edge: from rare metal to finished super‑radar
Control of the gallium supply chain
GaN starts with gallium, a soft metal produced largely as a by‑product of aluminium and zinc refining. China dominates global gallium production and in recent years has imposed export restrictions, particularly toward certain defence and high-tech users abroad.
This new heat-management technique slots neatly into that strategic picture. If China can pair control over gallium with a performance lead in GaN device engineering, it tightens its grip on a crucial class of “third‑generation” semiconductors used in everything from radars to power electronics.
The Xidian team positions its work as a stepping stone toward “fourth‑generation” materials such as gallium oxide, which promise even higher voltage handling and temperature tolerance but are still experimental. The know‑how gained in managing thermal interfaces today will likely matter even more with those tougher, hotter-running materials.
| Aspect | Traditional GaN radar chips | New Xidian approach |
|---|---|---|
| Bonding layer structure | Disordered micro‑islands | Smooth, uniform interface |
| Thermal resistance | Higher, worsens with use | Lower by roughly one third |
| Radar performance | Capped by heat build‑up | About 40% higher at same size and power |
| Cooling demands | Bulky systems for top-tier arrays | Potential for lighter, simpler cooling |
Beyond missiles and stealth jets: civilian spillovers
Satcom, 5G and 6G stand to gain
GaN power amplifiers do not live only on fighter noses or missile batteries. They also sit in satellite communication payloads, ground terminals and base stations for high-frequency 5G links, especially in Ka band.
Higher efficiency and better thermal behaviour can extend satellite lifetimes, since less energy gets lost as heat in orbit. On the ground, operators could reach the same coverage with fewer base stations or lower electricity bills, a rare combination in telecom engineering.
China has already been testing more exotic GaN-based devices. At the end of 2025, another Xidian team presented a prototype that converts ambient electromagnetic waves into usable electricity. That kind of work hints at broader ambitions in radio-frequency energy management, straddling communications, sensing and power harvesting.
The same family of chips that helps a fighter jet track targets could later power dense urban 6G networks or quietly recharge sensors from background radio noise.
What this means for radar competition
Scenario: a cooler, sharper air picture over the Western Pacific
Imagine a Chinese stealth fighter flying over the Western Pacific on a long patrol. With more thermally efficient GaN modules, its radar can run a more aggressive tracking pattern for longer without overheating. That allows the pilot to maintain a detailed air picture while still controlling emissions to reduce detectability.
On the other side, a warship relying on older-generation radar may struggle to match that range and refresh rate without significant cooling upgrades. Over hundreds of sorties and deployments, those small percentage gains accumulate into better situational awareness and more comfortable margins in a crisis.
Thermal headroom can also be traded for reliability. A radar designed to operate well below its new temperature limit can run for years with lower failure rates, easing maintenance for air forces or navies.
Key terms worth unpacking
GaN, bandgap and why heat hurts
Gallium nitride is called a “wide bandgap” semiconductor. The bandgap is the energy difference between electron states in the material. A wider gap lets devices handle higher voltages and temperatures, and operate at higher frequencies – great news for radar and power conversion.
The catch is that wide-bandgap devices tend to concentrate power in a smaller active area, which leads to steep local temperature rises. If heat cannot escape quickly through the underlying layers, the device’s performance droops or it breaks down.
This is why a subtle change deep inside the chip, at the interface between materials, can matter as much as visible features like antenna size or radar waveform design.
Benefits and risks on the strategic front
For China, success here brings several benefits: more capable radars across air, land, sea and space forces; a stronger export offering for partners buying Chinese defence electronics; and extra leverage in technology negotiations where access to advanced semiconductors is at stake.
For rivals, the risk is a widening performance gap in sensors that underpin missile defence, air policing and electronic warfare. Western labs are also pushing GaN hard, yet this specific approach to taming heat at the bonding layer signals that Beijing is determined to turn its materials advantage into fielded systems.
As with any lab result, questions remain: how reproducible the process is at industrial scale, how these chips behave after years of thermal cycling, and how quickly they can be certified for flight or space use. Those details will decide whether this stays a journal headline or becomes standard kit in the next generation of Chinese super‑radars.