China’s GaN Cooling Breakthrough Could Power Super-Radar Systems

As next-generation radars become more powerful and accurate, engineers face an adversary quieter than hostile aircraft: heat.

Chinese researchers say they have overcome a thermal constraint that has held back some of the world’s most sophisticated radars. The development could give Beijing an early advantage in “super-radar” systems for military and civilian applications.

How heat silently limits powerful radars

The most advanced modern radars seldom reach their limit because they cannot “see” far enough. More often, overheating stops them first.

An active electronically scanned array (AESA) radar – the flat, panel-style antenna fitted to stealth aircraft and advanced warships – contains thousands of miniature transmit/receive modules that generate pulses of radio energy. Every module is built around a semiconductor chip required to manage extreme power levels at very high frequencies.

For the past two decades, gallium nitride (GaN) has been the leading material for these chips. In comparison with older gallium arsenide technology, GaN operates at higher voltages and frequencies, while supporting far greater power density. This is why GaN is used prominently in radars for China’s J-20 and J-35 fighters, and increasingly features in US plans for upgraded F-35 radar systems.

There is, however, a drawback to that capability: GaN becomes very hot. Extremely hot.

When engineers raise power levels in important radar ranges including X-band and Ka-band, heat builds up in the chip more quickly than the cooling system can remove it. Beyond a certain threshold, increasing power no longer delivers better range or resolution. Instead, the device reaches a thermal wall.

This thermal bottleneck, more than clever circuit design, has quietly capped radar performance for two decades.

The “invisible layer” holding back GaN radar progress

The Chinese advance, announced in January 2026 by a team at Xidian University, does not alter the fundamental radar architecture. Its focus is instead an almost unseen chip component that is little known outside specialist circles: the bonding layer.

This exceptionally thin interface joins different semiconductor materials within a GaN power device. It has traditionally been made from aluminium nitride. As it grows, however, the layer develops microscopic islands and uneven formations. Rather than passing heat cleanly to the substrate and eventually the cooling system, these flaws retain it.

As a device runs, thermal resistance across this irregular interface gradually rises and efficiency falls. Once a particular temperature is reached, the chip cannot safely process additional power, even where the external cooling arrangement appears oversized on paper.

From thermal choke point to “heat highway”

The team led by researcher Zhou Hong says it has made the bonding layer grow in a far more consistent and ordered form. Put simply, it has changed a disordered thermal choke point into a smooth “heat highway”.

Data published by the group in the journal Science Advances reports:

• around 30% reduction in thermal resistance inside the device
• about 40% increase in RF power performance
• no increase in chip size or electrical power consumption

A roughly 40% boost in radar output, without a larger chip or higher energy use, is a rare combination in high-power electronics.

Zhou emphasises that the device footprint remained unchanged. In aerospace applications, where each millimetre beneath a fighter’s radar nose cone is fiercely contested, that point is as significant as the headline figures.

What a 40% power gain could mean in combat

Radar capability does not rise in a linear fashion, although additional power gives designers several choices simultaneously. Chinese accounts of the research suggest that the 40% improvement may provide:

• greater detection range without increasing antenna size
• more precise distinction between closely positioned targets at long range
• improved resilience against electronic jamming and clutter
• quicker refresh rates for tracking manoeuvring threats

For a stealth fighter, more power at the same apparent output could enable the aircraft to “see” sooner while transmitting less frequently or in shorter pulses. That makes it less likely that opposing sensors will detect it and establish a targeting solution.

For land-based air-defence radars, improved thermal control offers broader coverage from the same truck-mounted system, potentially reducing the number of units needed to monitor a given area of sky.

On mobile platforms such as warships, the improvement creates a choice: greater performance with existing cooling loads, or unchanged performance using lighter and simpler cooling equipment, freeing weight and space for fuel or weapons.

Platform	Main benefit	Operational impact
Stealth fighter	More range at similar emissions	Earlier threat detection, improved survivability
Ground-based radar	Wider coverage per site	Fewer radars to protect a region
Naval radar	Higher power without heavier cooling	Better tracking in bad weather and dense clutter
Satcom / 5G base station	Lower power use for the same signal	Reduced operating costs, denser networks

A semiconductor advantage supported by raw materials

This technical progress rests on a strong industrial foundation. China is the world’s largest producer of gallium, the crucial element in GaN semiconductors. Beijing has already limited exports of certain gallium-related products to companies associated with US defence programmes.

Influence over this upstream supply chain could allow Chinese laboratories to move faster from laboratory processes on test wafers to mass-produced radar modules for ships, aircraft and missile batteries.

Xidian University says the new bonding-layer technique reinforces China’s standing in so-called third-generation semiconductors, including GaN and silicon carbide, while laying foundations for fourth-generation materials such as gallium oxide. These future devices are expected to support still higher power levels and present even more demanding thermal problems.

Thermal management is turning into a strategic lever, not just an engineering detail, in the contest over advanced sensors and communications.

Civilian uses: satellites to 6G

The military is likely to be the first major customer for cooler and more powerful GaN chips. Nevertheless, the same underlying components are already used in civilian equipment.

In satellite communications, particularly in the Ka band, GaN power amplifiers strengthen transmissions between ground stations and orbit. More efficient heat removal could raise data throughput from a given satellite or expand coverage without launching additional spacecraft.

At ground level, 5G base stations and future 6G networks depend on dense assemblies of RF amplifiers. Better chip-level thermal performance reduces electricity costs and makes it easier to install powerful equipment at restricted urban locations, on towers and even on vehicles.

Researchers at Xidian have also worked on devices designed to turn stray electromagnetic waves into usable electricity, suggesting a wider strategy: extracting more value from every watt transmitted into, or captured from, the air.

What “super-radar” actually means

“Super-radar” is not an official technical term. It describes several trends converging: broader bandwidths, agile waveforms, higher output power, more intelligent signal processing and, crucially, components able to withstand severe heat over extended periods.

In practical terms, a super-radar could instantly move between long-range search, close target tracking, missile guidance and data-link functions through a single antenna face. It could also generate detailed synthetic-aperture radar imagery while continuing air-defence surveillance in the background.

Such functions are impossible if the hardware must reduce output every few seconds to cool down. This is the significance of the Chinese bonding-layer research: it addresses the limiting factor that often receives less attention than eye-catching algorithms or new waveforms.

Risks, compromises and the next stage

As with every new semiconductor manufacturing method, uncertainties remain. Production yields for complex GaN structures are notoriously difficult to manage. Creating a uniform bonding layer in laboratory conditions is one challenge; reproducing it across thousands of wafers at an industrial fabrication plant, at acceptable cost, is another.

Greater power density also raises the consequences of failure. A radar module operating at high temperatures can deteriorate rapidly or fail catastrophically if its thermal path is suddenly lost. Before approving fleet-wide deployment, defence customers will demand lengthy burn-in testing, severe environmental trials and extensive modelling.

For Western defence planners, the possibility that China could deploy radar arrays with 30–40% more effective power affects several areas at once. Stealth aircraft may encounter denser and more capable integrated air-defence networks. Chinese maritime surveillance from coastal areas could extend further into contested waters. Electronic warfare units could require improvements earlier than anticipated.

Yet the underlying physics does not belong to a single nation. Laboratories in the US, Europe, Japan and South Korea are all investigating improved thermal interfaces for GaN and related semiconductors. China’s announcement indicates that the competition is speeding up, and that the next major improvements in radar and communications could come not from software, but from a handful of atomic layers with better-managed heat.