A Novel Fanless Energy Efficient Edge Computing System Architecture and Engineering Practice for Baidu PCDN Application

The success of the Internet of Things (IoT) and rich cloud services have helped create the need for edge computing, in which data processing occurs in part at the network edge, rather than completely in the cloud. Edge computing targets to address concerns such as latency, mobile IoT devices' limited battery life, bandwidth costs, security, privacy and so on. Extending potential cloud applications beyond data center at the networking edge is one of hot technical topics among global cloud service providers (CSPs) nowdays. P2SP CDN (aka PCDN), as the first Baidu deployment quality cloud edge computing application, Baidu innovated put their content delivery network (CDN) application to networking edge for bandwidth utilization and cost efficiency, also accelerates end customer response to address the latency requirement of the emerging applications. Compared with conventional data center environment, PCDN environment is more complicated for infrastructure devices maintained by multiple parties, such as dynamical temperature, high humidity, corrosion prone, power constraint, non-stand-ardized management interface and so on. This paper proposes a novel fanless box system architecture, adapting for Baidu PCDN edge computing environment. Then details the comparison of conventional server system for energy and cost efficiency from system thermal and power architecture perspective. Engineering practice result and best known methods are summarized in the end. This paper also introduces Baidu PCDN application architecture and the methodology of system energy efficiency optimization based on workload characteristic.