Abstract
As the silicon-based CMOS technology faces significant challenges in terms of physical principles and process limitations by reducing the device dimensions to improve integration, the chiplet heterogeneous integration (CHI) system has become an important solution to continue Moore's law. The wafer-scale CHI systems have also emerged nowadays to achieve more powerful performance for artificial intelligence (AI) applications. Due to the high power and small space, the thermal problems are big challenges. Those problems can lead to severe damages if they are not well solved. In this paper, an efficient thermal model is proposed to predict the steady-state temperature distribution of CHI 2.5-D systems. Specifically, since an oversimplified heat sink can result in huge inaccuracy, the heat sink is modeled according to its actual shape. In addition, the RDL, TSVs and bumps are modeled in an equivalent manner to make this model more accurate. Those efforts enable steady-state thermal simulation of the CHI system. In order to extend the thermal model to wafer-scale CHI systems, an efficient method is developed to transform the circular components into regular shape. Finally, this thermal model is evaluated to be highly accurate with errors <1.8 % and quite fast. Based on this verification, several cases are analyzed to illustrate the applicability of our methodology in describing the thermal dissipation behaviors. The present thermal model is an instructive simulation tool which can be used to perform large scale thermal simulation of heterogeneous integration systems. The proposed methodology of the present simulation work is expected to help the designers to detect the potential temperature-dependent hotspots and improve the reliability and robustness of CHI systems in the early-stage of the design flow.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.