In the era of intelligence, everything digital and intelligent is driving ubiquitous computing, with a new business model centered on data accelerating the extension of computing power from data centers to the edge. Inspur Information believes that as edge computing further integrates with technologies such as artificial intelligence, cloud computing, the Internet of Things, and 5G, the edge computing infrastructure will become the intersection of IT, CT, DT, and industry OT, allowing computing to reach the boundaries of digital perception. E2X (Edge to everything) will create a “super neural network” for the future world of digital-physical integration. Therefore, Inspur Information is committed to building an edge computing “foundation” that meets the diverse and fragmented scenario demands while promoting the integration of the edge industry ecosystem through open standards, constructing a “product leads diversity, standards promote integration” edge computing + product and ecosystem system to drive the realization of E2X.
Diversity at the Edge, Standards First
Although edge computing is receiving increasing attention and action from many enterprises, significant challenges remain in practice. There exist numerous types of devices, protocols, and applications in various production environments, making it difficult to unify device interfaces, with systems operating independently and data hard to interconnect… Edge computing is facing multiple challenges from diverse scenarios, technologies, and ecosystems.
First, edge scenarios are rapidly landing, with new demands for the deployment of edge computing product solutions arising from seven major fields: ICT integration, next-generation CDN, intelligent connected vehicles, industrial internet, industry + AI, urban governance, and smart IoT. Second, as a technology empowerment platform, edge computing will naturally integrate deeply with technologies such as AI, big data, blockchain, and cloud, necessitating that edge computing products meet the growth characteristics of diverse technologies. Lastly, edge computing faces the challenge of integrating multiple ecosystems across IT, CT, and OT. For example, the diversity of protocol interfaces in the OT field leads to issues of interoperability and integration of industrial buses, while different communication protocols in the CT field create data synchronization issues.
Tian Hui, Director of the Integration Innovation Research Center of the China Academy of Information and Communications Technology, stated: “Edge computing encompasses numerous application fields such as mobile internet, IoT, and vehicle networking. The connectivity and data interaction of devices at the network edge are more complex, and standardization will help improve the compatibility and interoperability of edge computing systems, accelerating the popularization of edge computing technology, which will play an important role in promoting technological innovation and supporting the further development of the edge computing industry.”
Inspur Information has become a pioneer in edge computing industry standards, actively participating in and promoting the “Edge Computing +” project of the Computing and Networking Integration Industry and Standard Promotion Committee (CCSA TC621), proposing a referenceable overall architecture for “Edge Computing +” and relevant technical indicators and requirements for the integration of edge computing with 5G, artificial intelligence, audio and video, blockchain, security, and high-performance computing, building a “Edge Computing +” technological innovation system to meet industry application needs in efficient computing power, massive access, intelligent analysis, and security protection, and participating in the construction of overall architecture and requirement standards for edge computing aimed at the power industry, promoting the empowerment of edge computing in the power sector.
Additionally, Inspur Information is actively participating in promoting the ODCC Open Computing Organization’s OTII project for the telecommunications industry, working on the standard for a lightweight hardware architecture platform OTII that meets future telecommunications applications, and has released the industry’s first edge computing server that meets OTII standards, which has already been fully deployed in various scenarios such as ICT integration, smart oil wells, intelligent manufacturing, and smart transportation, accelerating the industrialization of innovative edge computing technologies. While continuously promoting the iteration and update of OTII standards, Inspur Information is also participating in and leading the definition of the edge microserver design standard ECOM construction, focusing on system architecture, and achieving rapid and flexible edge computing product design and development through the decoupling of functional modules such as computing, storage, IO, networking, and management.
Releasing Full-Stack Edge Computing Products to Meet Diverse Needs
Sun Bo, General Manager of the Edge Computing Product Department at Inspur Information, stated: “In addition to promoting the construction of the edge computing standards system, we should also focus on application needs and center on system design, achieving high coordination between software and hardware through product innovation to provide users with diverse forms, agile deployment, intelligent operation and maintenance, and strong environmental adaptability full-stack edge computing infrastructure that meets the complex and diverse scenario needs at the edge.”
As one of the earliest providers of computing infrastructure for edge computing, Inspur Information made a significant announcement at the conference, releasing the “4+3+X” full-stack product family for edge computing, including four major hardware product series for edge computing, three major software stacks, and X types of edge computing solutions such as cloud-edge collaboration, edge AI, hyper-convergence, and edge networking.
Based on its leading advantages and innovative capabilities in server architecture design, stringent heat dissipation, and high-end manufacturing, Inspur Information has built a hardware product array covering edge micro centers, edge servers, portable AI servers, and edge micro servers, along with three edge computing software platforms: edge vision, edge brain, and edge intelligence, to meet the diverse business needs of edge computing in various scenarios and environments, achieving collaborative optimization of the software and hardware systems of edge servers.
At this conference, Inspur Information launched two new generation G7 edge computing servers — the new flagship 2U dual-socket edge server NE5260G7 and the 1U single-socket edge server NE3160G7. Both products are designed based on the latest open computing OTII standards, with strong environmental adaptability and compact size. Among them, the NE5260G7 has significantly improved performance by 82% compared to the previous generation, while also offering a flexible product architecture that can support over ten types of acceleration chips for edge AI computing scenarios and can double the storage capacity without changing the physical size, providing up to hundreds of TB of storage capacity to fully meet the diverse needs of edge scenarios in AI computing, storage, and networking; while the NE3160G7 is a lightweight flagship edge server aimed at small base station scenarios, supporting three types of clock synchronization with a latency as low as microseconds, capable of stable operation in complex environments from -5 to 55℃. Additionally, the NE3160G7 features flexible network and AI acceleration card expansion, allowing for flexible configuration for different scenarios.
At the same time, the three edge computing software platforms: edge vision, edge brain, and edge intelligence can provide remote automated operation and maintenance management for edge server devices, intelligent management and scheduling of edge computing resources, and configuration management of edge AI algorithms, achieving a 95% increase in asset management efficiency and over 90% accuracy in predictive maintenance.
Furthermore, Inspur Information will collaborate with numerous partners for joint innovation to provide various edge computing scenario solutions to customers, achieving personalized and full-stack delivery of E2X scenarios such as edge cloud, far-edge, and near-edge.
Establishing an Edge Open Lab to Promote Industrialization
The sustainable development of edge computing requires continuous enhancement of integration and interoperability between different technologies and platforms; ecosystem construction is of utmost importance. At this conference, Inspur Information, together with Intel, Baidu, and 38 other partners, co-built an open edge integration lab and released the latest achievements aimed at industries such as telecommunications, transportation, water conservancy, and energy.
Dr. Zhang Yu, Senior Chief AI Engineer and CTO of the Network and Edge Division in China at Intel, believes that the establishment of the integration lab will build an “innovation consortium” for collaborative innovation in edge computing, promoting forward-looking technology research, scenario application validation, ecosystem resource integration, and innovative result incubation in telecommunications, manufacturing, energy, transportation, water conservancy, finance, ICT, and logistics, facilitating the healthy development of the edge computing industry.
Additionally, at this conference, Inspur Information and Baidu launched the roadside edge computing unit for high-level automated driving at urban intersections, which will provide roadside artificial intelligence computing power support for the first test section in Beijing that supports L2 to L4 automated driving.
The roadside edge computing unit is aimed at L2 to L4 high-level automated driving vehicle-road collaboration scenarios, capable of comprehensively sensing the status of traffic lights, lidar, road signs, weather stations, etc., at a two-way 8-lane intersection, providing computing power support for high-level automated driving, smart cities, smart transportation, and other applications, offering more precise real-time detection and analysis of all elements such as people, vehicles, roads, environments, and traffic events for “smart vehicles”, achieving smart transportation services through vehicle-road-cloud collaboration. Currently, the “perception-computation-communication” roadside edge intelligent system based on this core computing unit is undergoing deployment testing in Beijing’s high-level automated driving demonstration area, and will provide services for establishing a city-level “vehicle-road-cloud integrated” demonstration application area in Beijing in the future.
In the future, under the guidance of the “smart computing” strategy, Inspur Information will empower digital business innovation across various industries through continuous product innovation and ecosystem collaboration, with full-stack products and solutions for edge computing based on scenario innovation, accelerating digital transformation.
Source: Guangming Net