In a monumental development at the crossroads of semiconductor physics and quantum computing, researchers from the University of Hong Kong’s Department of Electrical and Computer Engineering and the Centre for Advanced Semiconductors and Integrated Circuits (CASIC) have unveiled a revolutionary programmable neuromorphic hardware platform that operates effectively at cryogenic temperatures near absolute zero. This breakthrough promises to resolve critical challenges impeding the scalability and efficiency of quantum computing systems, while simultaneously opening new frontiers in deep-space exploration technology.
At the heart of this innovation lies a novel exploitation of Silicon Carbide (SiC) MOSFETs, a semiconductor device that has traditionally seen use in power electronics and electric vehicle applications. Under extreme cryogenic conditions—temperatures plunging as low as 10 millikelvin—these devices exhibit a unique “S-shaped” negative differential resistance (NDR) behavior, an effect previously overlooked at such low-temperature regimes. This phenomenon is directly attributed to electron-donor impact ionization (EDII), an intrinsic atomic-scale mechanism that enables precise and stable modulation of current flow without reliance on thermal effects.
Leading the charge, Professor Yuhao Zhang and his PhD student Xin Yang pioneered the harnessing of this NDR in SiC MOSFETs to emulate the spiking behavior characteristic of biological neurons. Spiking neurons form the fundamental computational units in neuromorphic systems, mimicking the energy-efficient information processing of the human brain. Demonstrated for the first time at cryogenic temperatures, this transistor-level functionality signifies a colossal leap toward integrating neuromorphic architectures directly within the inherently cold environment of quantum processors.
Quantum computers operate by manipulating qubits, quantum analogs of classical bits, whose coherence is exquisitely sensitive to temperature fluctuations and environmental noise. Present control electronics rely predominantly on silicon-based devices operating at higher temperatures, which generate significant thermal loads and power consumption. Consequently, these electronics are spatially separated from qubits by long wiring connections, introducing latency and signal degradation that hamper system scalability.
The newly developed SiC neuromorphic platform, by functioning at millikelvin temperatures alongside qubits, drastically cuts down the wiring complexity and thermal footprint. The key lies in the material’s unique carrier dynamics under cryogenic conditions, enabling circuits that are thousands of times more energy-efficient compared to conventional silicon electronics. Such energy efficiency is critical in mitigating heat dissipation within dilution refrigerators—the cooling systems that maintain quantum processors at near absolute zero.
What distinguishes this approach further is the reproducibility and robustness of the NDR effect, which does not hinge on thermally activated processes but on purely electronic transitions governed by the atomic lattice configuration of SiC. This stability promises uniformity across manufacturing batches, a vital consideration for scalable chip production. Moreover, leveraging the well-established industrial infrastructure for SiC fabrication—primarily driven by electric vehicle and power grid applications—the team projects seamless industrial-scale manufacturing of these cryogenic devices on 300-mm wafers.
The researchers demonstrated that individual neuromorphic transistors can be interconnected to form cascading networks capable of complex data processing within the cryogenic environment. Such local processing could greatly enhance quantum error correction mechanisms—a cornerstone requirement for fault-tolerant quantum computing—by performing rapid, real-time control operations without transferring signals out of the ultra-cold domain.
Beyond quantum computing, the ruggedness and low-temperature operability of these SiC-based neuromorphic circuits make them exceptionally suitable for extraterrestrial missions. Electronics deployed on the lunar surface or deep space probes must endure extreme temperature fluctuations, often dipping into cryogenic ranges, where traditional silicon-based devices falter. This technology paves the way for intelligent onboard systems capable of autonomous decision-making and adaptive controls far from Earth.
Published in the prestigious journal Nature Communications, the study titled “Cryogenic neuromorphic circuits using gate-controlled negative differential resistance in silicon carbide” meticulously details the experimental investigations, device physics, and circuit architectures underlying this technological feat. The article is accessible via the digital object identifier 10.1038/s41467-026-70963-6.
Professor Yuhao Zhang, whose career spans credentialed tenure at MIT and leadership roles at Virginia Tech before joining HKU, spearheaded this interdisciplinary endeavor blending power electronics, materials science, and quantum engineering. His extensive portfolio includes over 200 scholarly articles, multiple patents, and recognition from premier institutions such as IEEE and the National Science Foundation.
By bridging the expansive gap between semiconductor physics and quantum information science, this breakthrough stands to redefine the architectural paradigms for next-generation quantum computers. The convergence of cryogenic neuromorphic hardware with quantum processors introduces a new paradigm in scalable, energy-efficient quantum control frameworks—ushering in an era where quantum technologies transition from experimental novelties to practical, robust systems.
As the quantum computing landscape continues to mature, innovations like this programmable SiC neural platform not only propel computational capacities but also stimulate synergies across allied fields, including robotics, space exploration, and advanced AI hardware. The research marks a seminal milestone that is poised to reverberate through the scientific and technological community, catalyzing subsequent explorations into the untapped potential of cryogenic electronics.
This discovery, combining deep theoretical insight with pragmatic industrial feasibility, underscores the critical role of novel materials and device physics in overcoming bottlenecks inherent to ultra-low temperature quantum systems. Future work envisages integration of these neuromorphic chips within fully operational quantum computing hardware, charting the course for unprecedented computational throughput enveloped in the quantum regime.
The confluence of established SiC fabrication technology and groundbreaking cryogenic operation heralds a new era, one where quantum machines achieve true scalability, and space exploration hardware attains unfathomable resilience, enabling humankind to grasp the farthest frontiers with sophisticated local computation embedded within the cold embrace of the cosmos.
Subject of Research: Not applicable
Article Title: Cryogenic neuromorphic circuits using gate-controlled negative differential resistance in silicon carbide
News Publication Date: 23-Mar-2026
Web References: https://www.nature.com/articles/s41467-026-70963-6
References: 10.1038/s41467-026-70963-6
Image Credits: Not provided
Keywords
Applied sciences and engineering, Engineering, Electrical engineering, Robotics, Quantum computing, Cryogenic electronics, Silicon Carbide MOSFETs, Neuromorphic hardware, Negative differential resistance, Quantum error correction, Deep-space exploration, Cryogenic neuromorphic circuits
