Embedded Core-Based Machine Learning Design for Baseboard Management Controller

The advent of the Fourth Industrial Revolution has spurred active research and business development across diverse domains, driven by the escalating demands for big data, AI, and cloud services. As industries progress, the demand for data centers has surged significantly, with a heightened emphasis on curbing power consumption within these facilities. This study delves into the design of On-Device AI for reducing power consumption, within the context of Baseboard Management System (BMC) firmware. In this research, the BMC under development is endowed with On-Device AI functionality, aimed at both power consumption reduction and the automation of data server failure and malfunction management. The principal objective of this study is to formulate a BMC tailored for data servers, aligned with the prevailing trends of the Fourth Industrial Revolution era, and centered on power efficiency enhancements as well as automated server failure and malfunction management.