A novel NVM memory file system for edge intelligence

Edge intelligence (EI), as a combination of artificial (AI) and internet things (IoT), is the key to realize an interconnected world everything. EI needs process data collected by edge devices locally. For with insufficient computing power, near (NDC) system can provide local capability adding co-processing unit memory. NDC systems are extensively studied focused on optimization hardware architecture but lack universal transparent support. Meanwhile, we notice that non-volatile memory (NVM) plays important role in intermittent computing. In this article, propose novel in-memory processing file (IMPFS) based NVM general support for running devices. IMPFS uses standard interfaces optimized AI application management. To verify IMPFS, prototype designed. The results showed not only also further improve speed reducing redundant software overhead.