At NVIDIA GTC21, Inspur Information announced the launch of its Intelligent Edge Microserver EIS200 with compatibility and optimization for the newly released NVIDIA Jetson TX2 NX AI Edge Inference System on Module (SOM) in addition to existing compatibility with multiple AI edge modules like the Jetson Nano and Xavier NX SOMs. This helps deliver a versatile, lightweight edge AI computing selection for a broad range of scenarios, including intelligent oil well, power patrol, industrial quality control (QC), intelligent transportation and intelligent retail. Inspur’s Intelligent Edge Microserver EIS200 features a compact form factor and strong AI performance ranging from 0.5 Tera operations per second (TOPS) to 21 TOPS to meet diverse computing power requirements. The new Jetson TX2 NX-based EIS 200 supports 1080p video encoding and decoding for up to 14 channels. Inspur EIS-200 series now provides a 144% price-performance ratio improvement compared to Jetson Nano-based setups, supporting a broad range of specific edge computing applications for image, video and voice analysis with only 25W power consumption. Use cases include AI QC and analysis of industrial components and parts, fast identification of moving items at crossroads, human gesture recognition for production safety and protection, and intelligent identification and control of oil well operations. The EIS200 server was benchmarked under the classic network ResNet50 which is based on classification, where the performance reached 112 frames per second (FPS), which is 2.24 times that of Jetson Nano-based setups. When it was benchmarked under the classic network tiny YOLO-V3 which is based on object detection, the performance reached 108 FPS, which is 2.25 times that of Jetson Nano-based setups. This gives developers and manufacturers faster inference compute speeds with up to 50% decrease in inference time needed. As near-edge devices get closer to the sensing layer of scenarios, they need to undertake mass data interactions and decision making. The solutions to problems like the adaptability of computing devices to the ambient environment as well as data transmission and large-scale deployment and O&M in different scenarios have become the key to the rapid development of edge computing and AI-driven businesses. The Intelligent Edge Microserver EIS200 has taken into full account the demanding deployment environment in edge computing scenarios and enabled a stable performance at the temperature of -40 to 60 Celsius degrees through fanless aluminum-case cooling under an IP40 industrial-level protection. It supports comprehensive radio communication modalities ranging from short range to long range via its extensive interfaces. In businesses that need mass deployment, the Intelligent Edge Microserver EIS200 has also realized minimalist O&M for a uniform deployment and update of inference algorithms via the data center-level management platform with significantly improved device management efficiency. Inspur has diversified its edge computing products in multiple forms and collaborated with partners from both the upstream and downstream of the industrial chains to develop scenario-based full stack solutions for seven major edge computing scenarios such as industrial Internet, intelligent connected vehicles, urban governance, and industry + AI. The Jetson TX2 NX-based Intelligent Edge Microserver EIS200, with its core advantages like super strong computing performance, open ecosystem, and minimalist O&M, will further advance the industrial application of AI in edge computing scenarios and drive the digital and intelligent transformation in all industries, including manufacturing, transportation, energy, finance and healthcare.