A Thermal Profile Prediction Methodology for Nanosheet Circuits Featuring Cross-Layer Thermal Coupling Effect

The thermal coupling within gate-all-around nanosheet (GAA-NS) circuit cannot be neglected as technology node scales down. In this work, a thermal profile prediction methodology is proposed to accurately and efficiently assess the thermal effect of GAA-NS circuits. By comprehensively considering the thermal coupling effect in devices and interconnects, the cross-layer thermal resistance network model is established. The accuracy of the model is well verified by comparing with finite element simulation in device level and circuit level. Based on this model, the thermal impact on circuit performance and reliability can be predicted. The results show that an evident error will be introduced if not considering these thermal coupling effect. The calculation speed of this model is 23 times that of the finite element method, and the calculation efficiency is greatly improved while the calculation accuracy is maintained. Thus, this work is helpful to the resilient thermal-aware circuit design beyond 3nm node.