In a groundbreaking study led by Professor Yuhao Zhang and PhD student Xin Yang, researchers have unveiled a novel approach for generating and controlling negative differential resistance (NDR) within industry-standard Silicon Carbide (SiC) MOSFETs. This innovative method allows a single transistor to mimic the energy-efficient "spiking" behavior of biological neurons at extraordinarily low temperatures, as low as 10mK.
Transforming Quantum Control Electronics
Quantum computing relies on complex control electronics to manage qubits, which need to be maintained at millikelvin temperatures. Traditional silicon-based control systems often consume excessive power and produce unwanted heat, necessitating their placement away from the qubits. This spatial separation creates extensive wiring needs, complicating the construction of large-scale quantum computers.
"Our research presents a hardware platform that can be seamlessly integrated with quantum processors," Professor Zhang explained. "Utilizing the unique carrier dynamics of silicon carbide, we can produce circuits that are thousands of times more energy-efficient than standard electronics, significantly alleviating the thermal burden on cryogenic systems."
Unique Cryogenic Properties of Silicon Carbide
The research team discovered that SiC MOSFETs exhibit a pronounced "S-shape" NDR effect when cooled below 2K, a phenomenon driven by electron-donor impact ionization (EDII). Unlike other technologies that generate heat within the device, this newly identified mechanism stems from the atomic characteristics of the material, ensuring high stability and consistent reproducibility across various manufacturing batches.
"This approach is both robust and scalable," Mr. Yang noted. "Since SiC is already widely utilized in electric vehicles and power grids, we can harness existing industrial foundries to produce these cryogenic chips on 300-mm wafers."
Applications Beyond Quantum Computing
The study also showcased the ability to connect these artificial neurons into larger networks, paving the way for advanced local data processing at cryogenic temperatures. This could enhance critical quantum computing functions such as quantum error correction and real-time quantum control.
The implications of this technology extend far beyond quantum computing. Designed to operate reliably in extremely cold environments, these circuits could prove invaluable for deep space exploration. Future systems may function effectively in the harsh conditions of the Moon's surface or the more distant reaches of our solar system.
The findings were published in Nature Communications under the title "Cryogenic neuromorphic circuits using gate-controlled negative differential resistance in silicon carbide."