The rapid proliferation of artificial intelligence (AI) servers and high-performance computing systems has significantly elevated the technical and reliability requirements for multilayer ceramic capacitors (MLCCs). In such systems, MLCCs are critical passive components that must deliver high capacitance, fast transient response, and robust insulation performance under high temperature, voltage, and current density. This review examines the material, structural, and process innovations that underpin MLCC performance in AI applications. Key topics include the development of ultrathin dielectric layers (<0.5 μm), rare-earth doped BaTiO₃-based dielectrics with enhanced DC bias stability, and core-shell microstructures designed for temperature and field resilience. The paper also explores insulation degradation mechanisms―such as vacancydriven conduction and demixing―and advanced reliability assessment methodologies, including HALT, TSDC, and the tipping point framework. Comparisons with automotive-grade MLCCs highlight the unique requirements of AI systems, such as ultraminiaturization, high volumetric efficiency, and ppm-level field failure rates. Finally, the review discusses emerging trends in MLCC technology, including particle engineering, interface stabilization, and advanced lamination techniques, and provides insight into the future direction of capacitor development tailored to AI data center environments.

In this study, we investigated the color change of the normal light gray granite as the high value color granite. By coating the metal catalyst liquid on the surface of granite stone, the metal particles were penetrated into the granite and the color of granite was changed permanently through the annealing treatment. To increase penetration depth into the granite, we used DC (direct current) bias. Two kinds of bias were used such as DC bias and pulse DC bias. And the penetration time was changed as 30 and 60 min. In all cases, the color granite were successfully obtained. Regardless of the catalyst reaction time, the penetration depth was increased by using the bias treatment. We obtained a penetration depth of 21 mm with the DC pulse bias during 60 min.

Hydrothermal synthesis technique could be carried out for growth of ZnO nanowires atrelatively low process temperature, and it could be freely utilized with various substrates for fabricationprocess of functional electronic devices. However, it has also a demerit of relatively slow growthcharacteristics of the resulting ZnO nanowires. In this paper, an external DC bias of positive and negative0.5 [V] was applied in the hydrothermal synthesis process for 2∼8 [h] to prepare ZnO nanowires on aseed layer of AZO with high electrical conductivity. Growth characteristics of the synthesized ZnOnanowires were analyzed by FE-SEM. Material property of the grown ZnO nanowires was examined byPL analysis. The ZnO nanowires grown with positive bias revealed distinctively enhanced growthcharacteristics, and they showed a typical material property of ZnO.