AI COMPUTING

Thanks to their high voltage capability, thermal robustness, and high frequency performance, wide bandgap semiconductors, especially Silicon Carbide (SiC) and Gallium Nitride (GaN), are essential to improving the energy efficiency of AI computing infrastructure.
BOOSTING ENERGY EFFICIENCY TO MEET HIGH POWER DEMAND
As AI workloads surge, data center power consumption has also skyrocketed. SiC and GaN power devices dramatically reduce conversion losses. In server power supplies (PSUs), SiC-based modules can achieve over 97% system efficiency, significantly lowering energy waste compared with traditional Silicon solutions.
HVDC ARCHITECTURE & ULTRA-HIGH POWER DENSITY
Power industry is shifting towards 800V high voltage DC (HVDC) architectures to address low efficiency power distribution when power exceeds 200 kW. Wide bandgap devices enable next generation solid-state transformer (SST) systems that convert 13.8 kV AC directly to 800 V DC. This can eliminate multiple conversion stages, improving transmission efficiency by 5%, and reducing copper usage by 45%.
Meanwhile, the high frequency capability of GaN devices allows much smaller magnetic components, reducing overall PSU size by 30 to 50%. This is ideal for spatially constrained data centers.
REDUCED COOLING REQUIREMENTS & HIGHER SYSTEM RELIABILITY
SiC’s thermal conductivity (three times that of silicon), combined with GaN’s low switching losses, can significantly reduce heat generation. This greatly decreases cooling demands, lowers data center PUE, and enhances long term system reliability.
In short, wide bandgap semiconductors improve efficiency, increase power density, and strengthen system reliability. They become the essential “power heart” behind sustainable AI compute growth and one of the key technologies for overcoming the energy barriers of high performance computing.