Every time a high-power radar lights up the sky, billions of tiny components heat up and fight for survival. Chinese researchers now claim they’ve cracked the thermal bottleneck holding back one of the most critical technologies in modern warfare.
How heat quietly caps the range of modern radars
On paper, radars are limited by physics: power, frequency, antenna size. In reality, they often hit a different wall first: the chip overheats long before engineers reach the theoretical range.
Most advanced military systems now rely on gallium nitride (GaN) semiconductors. GaN can handle higher voltages and faster switching than older gallium arsenide chips. That’s why it sits inside modern active electronically scanned array (AESA) radars on aircraft such as China’s J-20 and J-35, and why the US is pushing GaN modules into the F‑35’s radar upgrades.
But GaN has a notorious downside. As power ramps up, especially in X‑band and Ka‑band frequencies used for fire control, long‑range tracking and satellite links, the device heats up much faster than that heat can escape.
For two decades, thermal limits rather than transistor design have quietly set the ceiling on radar performance.
Turn the power up, and the range, resolution and refresh rate improve – until the chip starts to cook. Turn it down, and the radar lives longer, but sees less and reacts more slowly. This trade‑off has defined GaN radar design since the early 2000s.
The hidden layer that was choking GaN performance
The new Chinese work, led by researcher Zhou Hong at Xidian University, doesn’t change how a radar works. It changes how the heat gets out of the chip.
The interface that turned into a thermal trap
Inside a GaN power device, several semiconductor layers are stacked on top of each other. At the junction between these layers sits an ultra‑thin “buffer” or bonding layer, often made from aluminium nitride.
In current manufacturing processes, that layer tends to grow in microscopic, uneven “islands”. These irregular structures disrupt the path that heat is supposed to follow from the active region of the device down into the substrate, where it can be carried away.
➡️ Day will turn to night: the longest total solar eclipse of the century now has an official date
➡️ Eclipse of the century: 6 minutes of darkness: when it will happen and where to watch it
➡️ Light fast apple cake made with oil and yogurt for effortless desserts
The result: hot spots build up, thermal resistance rises as the device ages, and performance plateaus well below what the material should in theory allow.
According to Xidian, the team managed to force that interface layer to grow as a smooth, uniform film rather than a patchwork of islands. That effectively converts a narrow thermal bottleneck into a much wider channel.
- Thermal resistance cut by roughly one third
- Radar output performance boosted by about 40% for the same chip size and power draw
A 40% jump in usable power without a bigger chip or extra electricity reshapes the design rules for high‑end radars.
What does “40% more power” actually mean on a radar?
Engineers talk in watts and decibels, but users care about what appears on the screen. Extra power tends to translate into several concrete benefits:
- Longer detection range without enlarging the antenna
- Sharper target discrimination at long distance
- Better resilience against jamming and clutter
- Faster refresh when tracking fast or manoeuvring threats
For a stealth aircraft, extra radar punch without higher emissions is especially valuable. The aircraft can detect threats earlier while still keeping its own signature relatively low. For a ground‑based air‑defence radar, the same hardware can cover a larger volume of sky without heavier cooling systems.
Zhou’s team stresses that the gain came without widening the chip footprint, a critical factor for cramped fighter noses where every millimetre matters. The same thermal trick could, in principle, be applied to base stations, satellite payloads and even future 6G infrastructure, extending coverage while trimming power bills.
China’s advantage: materials, supply chain and timing
There is also a geopolitical angle. China already dominates global production of gallium, the key ingredient for GaN semiconductors. Beijing has tightened export controls on this strategic metal, targeting foreign defence users and sensitive tech suppliers.
By improving GaN device performance at the materials level, Chinese labs are adding more value on top of raw resource control. Xidian University describes the work as part of a broader push into so‑called “third‑generation” semiconductors, which include GaN and silicon carbide, and as a stepping stone toward “fourth‑generation” materials such as gallium oxide.
Control over both the raw gallium and the know‑how to turn it into superior radar chips gives Beijing leverage on two fronts at once.
That combination makes any breakthrough in GaN performance more than just an academic story. It touches export policy, allied defence planning, and the long‑running technology rivalry with the United States.
From fighter radars to 6G towers
Military systems are only the first in line
The earliest beneficiaries will almost certainly be military radars. High‑power GaN modules sit in airborne AESA radars, ground‑based air defence networks and shipborne surveillance systems. Every extra degree of thermal margin allows these systems to run harder, longer, or in more compact form factors.
But GaN power amplifiers are also the beating heart of many microwave communication systems. Ka‑band satellite links, advanced 5G base stations and experimental 6G testbeds all rely heavily on them.
Less thermal resistance means:
- More data throughput per antenna panel
- Smaller, lighter satellite payloads
- Lower cooling requirements at remote telecom sites
In December, another Xidian team showcased an experimental device that harvested usable electrical power from incoming electromagnetic waves. That type of work, combined with better thermal handling, hints at future systems that both communicate and scavenge energy at the same time.
Why this kind of heat management matters
For non‑specialists, “thermal resistance” can sound abstract. In practice, it’s similar to plumbing. If the pipe under your sink is too narrow or clogged, water backs up. In a GaN chip, the “water” is heat, and the narrow junction layer was the pipe that kept blocking.
By smoothing that hidden layer, the Chinese team effectively widened the pipe. The heat can exit the active region faster, so engineers can either raise the power or maintain the same power with a lower operating temperature, extending device life.
| Design choice | Without new interface | With improved interface |
|---|---|---|
| High power mode | Risk of hot spots, limited duty cycle | Higher power sustained for longer |
| Low power, long life | Conservative settings, wasted potential | Same output with less stress on the device |
| Compact designs | Extra cooling hardware needed | Smaller cooling systems possible |
In radar design, those margins cascade through the whole system. A lighter cooling unit frees weight for fuel or weapons. A smaller antenna nose lets designers tweak the aerodynamics. On satellites, a few kilograms saved on radiators can translate into lower launch costs.
Risks, limits and what could come next
This kind of advance does not remove all constraints. At very high powers, mechanical stresses from repeated heating and cooling cycles can still crack materials. Electromigration and radiation effects in space remain concerns. Any large‑scale military deployment would need exhaustive reliability testing across thousands of chips and many years of operation.
There is also the question of who else can replicate the process. If the growth method for the smooth interface layer depends on specific equipment or proprietary epitaxy recipes, Chinese firms might hold a lead for several years. If not, US, European, Japanese and South Korean players could adapt similar tricks in their own GaN foundries.
For defence planners, a simple scenario illustrates the stakes: picture a Chinese stealth fighter whose radar can track targets 20–30% farther out than analysts assumed, or a coastal missile battery covering a wider arc with the same number of launchers. That shifts timelines for detection, interception and escalation.
Outside defence, civilian telecoms and satellite operators may view this development through a different lens. Improved GaN thermal performance points to denser base stations, more capable phased‑array user terminals for broadband from orbit, and possibly more energy‑efficient infrastructure at a time when network power consumption is under scrutiny.
For readers trying to follow the jargon, two terms are worth keeping in mind: “GaN” simply refers to a semiconductor material much better suited than silicon for high‑frequency, high‑power work; “thermal resistance” describes how hard it is for heat to move through a structure. Lowering that resistance is turning into one of the quiet battlegrounds of both next‑generation radar and advanced wireless communications.