The rapid expansion of data-centric technologies has made energy-efficient computing a technological and societal priority. Modern systems consume energy not only through computation but also through data movement between logic and memory, resistive losses in interconnects and heat removal. These energy costs intensify as devices shrink and integration density increases. Architectural strategies such as three-dimensional integration reduce latency and increase functional density by bringing memory and logic closer together. Yet further scaling is progressively constrained by materials functionality at interfaces, rather than by bulk properties. In this Perspective, we argue that transport within the first few atomic layers at material boundaries will determine the efficiency and scalability of future electronic devices. First, we outline four governing principles of interface engineering: electrostatics, electronic hybridization, boundary-dominated transport and thermal and structural stability. We then derive design rules and use representative device examples to show how engineered interfaces control switching energy, electrical conduction, heat dissipation and reliability. Finally, we consider the need for low-temperature materials synthesis, interface-sensitive metrology and uncertainty-aware predictive modelling to facilitate future advances in energy-efficient computing. As electronic devices downscale, the energy efficiency of computing is increasingly limited by interfacial transport processes. This Perspective outlines how engineered electronic-material interfaces control switching, charge and heat transport, and reliability and discusses interface design for logic, memory, interconnects and thermal management, whereas identifying predictive modelling needs for interface design.
原文链接
分享
引用
