摘要
人工智能正在大跨步进入应用新时代:人工智能在大模型加持下性能大幅超越历史,而云原生边缘计算使得人工智能离所有用户更近。这两者的有机融合促使云原生边缘智能爆发新的价值增长点。本文基于历年实践,介绍了云原生边缘智能的核心变化,同时总结出最佳实践:(1)在指导理念方面,针对四大挑战解决实践问题,基于标准化范式推动多种技术突破;(2)在实现路径方面:以基准测试孵化算法,以统一框架开发服务,以生态社区发展产业。文章最后介绍了近十个场景二十项应用案例,并以开源社区生态Kube Edge SIG AI子项目标准化统一框架Kube Edge-Sedna和基准测试Kube Edge-Ianvs为基石,深入分享了云原生边缘智能产业应用经验。
Artificial Intelligence(Al)is making significant strides into a new era of application:Al performance has greatly exceeded history under the support of large-scale models,while cloud native edge computing has brought Al closer to all users.The integration of these two promotes the explosion of new value in cloud native edge intelligence.Based on years of practice,this paper introduces the core changes of cloud native edge intelligence and summarize the best practices:1)The guiding principle is to solve practical problems in response to four critical challenges,and promote various technological breakthroughs based on standardized paradigms;2)The implementation path is incubating algorithms with benchmark testing,developing services with a unified framework,and developing industries with ecological communitiesn conclusion,the article showcases nearly ten scenarios with twenty application cases,leveraging the KubeEdge SIG Al subprojects,namely the standardized unified framework of KubeEdge-Sedna and the benchmark testing tools of KubeEdge-lanvs,to profoundly share insights into the application of cloudnative edge intelligence in the industrial sector.
出处
《自动化博览》
2024年第2期53-58,共6页
Automation Panorama1
关键词
云原生边缘智能
人工智能
云原生边缘计算
Cloud-native edge intelligence
Artificial intelligence
Cloudnative edge computing
