Subjects: Other Disciplines >> Synthetic discipline submitted time 2023-03-28 Cooperative journals: 《中国科学院院刊》
Abstract： Independence and openness are equally important for the development of science and technology industries in an increasingly competitive international environment. Independence serves as our backup force, while openness allows us to participate fully in global division of labor and collaboration and to gain competitive edge. This paper analyzes the fundamental technology innovations in the history of computer industry: the mobile chip ARM and the mobile OS Android, as well as the developing intelligent computer industry, and explores how to build an applied science and technology system with competitive edge under the new international competitive environment. Two conclusions are drawn as follows. First, independence and full sharing of the benefits of international division of labor and cooperation must be based on the fundamental technology innovation starting from the source or backbone. Intelligent computer offers an opportunity for China to establish an independent and open industry. Second, only by changing the evaluative methodology and metric from simply centering on the citation number of papers to emphasizing fundamental technology innovation starting from the source or backbone, changing the top-heavy discipline layout, emphasizing intellectual property rights protection and making full use of international intelligence to carry out division and cooperation, can an applied science and technology system with internationally competitive edge be established.
Subjects: Computer Science >> Computer Application Technology submitted time 2016-11-15
Abstract：最新的研究分析表明，服务器购置费用仍然是影响大规模数据中心或云系统开销的主要因素。在本文中，我们认为经典的资源供应问题已经发生显著变化：异构负载已经成为大规模数据中心的常态，而目前的资源供应方法并不能很好地应对这种异构性。本文报道了我们针对这一问题展开的工作。我们的贡献有以下三个：首先，我们提出了一个协同资源供应的解决方案，通过充分利用异构负载的差异，降低在竞争条件下的尖峰负载的资源消耗；第二，基于四种典型的异构负载：并行批处理作业、Web 服务、搜索引擎和MapReduce作业，我们构建了支持异构负载的PhoenixCloud系统，能够有效地进行协同资源供应；第三，我们通过使用真实和合成的负载日志进行实验对系统进行了综合的评价。实验结果显示，我们的解决方案与现有的非协同解决方案相比可以极大节省服务器购置成本，该方案可以广泛用于当前的托管数据中心或云系统中，如利用统计复用技术的EC2 系统或实现了初步弹性资源供应技术的RightScale系统。
Peer Review Status:
Awaiting Review
Subjects: Computer Science >> Computer Application Technology submitted time 2016-05-04
Abstract：最新的研究分析表明，服务器购置费用仍然是影响大规模数据中心或云系统开销的主要因素。在本文中，我们认为经典的资源供应问题已经发生显著变化：异构负载已经成为大规模数据中心的常态，而目前的资源供应方法并不能很好地应对这种异构性。本文报道了我们针对这一问题展开的工作。我们的贡献有以下三个：首先，我们提出了一个协同资源供应的解决方案，通过充分利用异构负载的差异，降低在竞争条件下的尖峰负载的资源消耗；第二，基于四种典型的异构负载：并行批处理作业、Web 服务、搜索引擎和MapReduce 作业，我们构建了支持异构负载的PhoenixCloud 系统，能够有效地进行协同资源供应；第三，我们通过使用真实和合成的负载日志进行实验对系统进行了综合的评价。实验结果显示，我们的解决方案与现有的非协同解决方案相比可以极大节省服务器购置成本，该方案可以广泛用于当前的托管数据中心或云系统中，如利用统计复用技术的EC2 系统或实现了初步弹性资源供应技术的RightScale 系统。
Peer Review Status:Awaiting Review
Hits
Hits
Downloads
Comment