Interconnection network plays an important role in scalable high performance computer （HPC） systems. The TH Express-2 interconnect has been used in MilkyWay-2 system to provide high-bandwidth and low-latency interpr...Interconnection network plays an important role in scalable high performance computer （HPC） systems. The TH Express-2 interconnect has been used in MilkyWay-2 system to provide high-bandwidth and low-latency interprocessot communications, and continuous efforts are devoted to the development of our proprietary interconnect. This paper describes the state-of-the-art of our proprietary interconnect, especially emphasizing on the design of network interface. Several key features are introduced, such as user-level communication, remote direct memory access, offload collective operation, and hardware reliable end-to-end communication, etc. The design of a low level message passing infrastructures and an upper message passing services are also proposed. The preliminary performance results demonstrate the efficiency of the TH interconnect interface.展开更多
Due to the rapid development of the maritime networks, there has been a growing demand for computation-intensive applications which have various energy consumption, transmission bandwidth and computing latency require...Due to the rapid development of the maritime networks, there has been a growing demand for computation-intensive applications which have various energy consumption, transmission bandwidth and computing latency requirements. Mobile edge computing(MEC) can efficiently minimize computational latency by offloading computation tasks by the terrestrial access network. In this work, we introduce a space-air-ground-sea integrated network architecture with edge and cloud computing components to provide flexible hybrid computing service for maritime service. In the integrated network, satellites and unmanned aerial vehicles(UAVs) provide the users with edge computing services and network access. Based on the architecture, the joint communication and computation resource allocation problem is modelled as a complex decision process, and a deep reinforcement learning based solution is designed to solve the complex optimization problem. Finally, numerical results verify that the proposed approach can improve the communication and computing efficiency greatly.展开更多
With the increasing demand for data traffic and with the massive foreseen deployment of the Internet of Things (IoT), higher data rates and capacity are required in mobile networks. While Heterogeneous Networks (HetNe...With the increasing demand for data traffic and with the massive foreseen deployment of the Internet of Things (IoT), higher data rates and capacity are required in mobile networks. While Heterogeneous Networks (HetNets) are under study toward 5G technology, Wireless Fidelity (WiFi) Access Points (APs) are considered a potential layer within those multiple Radio Access Technologies (RATs). For this purpose, we have proposed in this paper a novel WiFi dimensioning method, to offload data traffic from Long Term Evolution (LTE) to WiFi, by transferring the LTE energy consuming heavy users, to the WiFi network. First, we have calculated the remaining available capacity of the WiFi network based on the estimated load of each WiFi physical channel using the overlapping characteristic of the channels. Then, we were able through this dimensioning method, to calculate the minimum needed number of WiFi APs that ensure the same or better throughput for the LTE transferred users. By this method, we have ensured additional capacity in the LTE network with minimum investment cost in the WiFi network. Finally, we have estimated the profit sharing between LTE and WiFi by considering data bundles subscription revenues and the infrastructure capital and operational costs. We have calculated for each network the profit share using a coalition game theory Shapley value that pinpoints the benefit of the cooperation using the proposed dimensioning method.展开更多
近十年来,随着网络技术的发展,网络带宽迅速增长,而同期CPU的性能未得到相应的提高。在吉比特网络下,网络终端CPU处理TCP/IP协议的能力已经成为限制网络应用的瓶颈。为了使终端用户能充分利用广阔的带宽资源,需要提高网络终端的协议处...近十年来,随着网络技术的发展,网络带宽迅速增长,而同期CPU的性能未得到相应的提高。在吉比特网络下,网络终端CPU处理TCP/IP协议的能力已经成为限制网络应用的瓶颈。为了使终端用户能充分利用广阔的带宽资源,需要提高网络终端的协议处理能力。文中基于FPGA的硬件设计,将原来由软件完成的IP层协议功能完全卸载出来,向CPU提供硬件支持。并且通过功能仿真、综合后仿真、布局布线后仿真验证了设计的可行性,由静态时序分析可知,协议处理器的时钟频率可达50 MH z,处理IP数据流的能力可以达到1.6 G b/s的网络线速度。展开更多
Digital earth science data originated from sensors aboard satellites and platforms such as airplane,UAV,and mobile systems are increasingly available with high spectral,spatial,vertical,and temporal resolution data.Wh...Digital earth science data originated from sensors aboard satellites and platforms such as airplane,UAV,and mobile systems are increasingly available with high spectral,spatial,vertical,and temporal resolution data.When such big earth science data are processed and analyzed via geocomputation solutions,or utilized in geospatial simulation or modeling,considerable computing power and resources are necessary to complete the tasks.While classic computer clusters equipped by central processing units(CPUs)and the new computing resources of graphics processing units(GPUs)have been deployed in handling big earth data,coprocessors based on the Intel’s Many Integrated Core(MIC)Architecture are emerging and adopted in many high-performance computer clusters.This paper introduces how to efficiently utilize Intel’s Xeon Phi multicore processors and MIC coprocessors for scalable geocomputation and geo-simulation by implementing two algorithms,Maximum Likelihood Classification(MLC)and Cellular Automata(CA),on supercomputer Beacon,a cluster of MICs.Four different programming models are examined,including(1)the native model,(2)the offload model,(3)the symmetric model,and(4)the hybrid-offload model.It can be concluded that while different kinds of parallel programming models can enable big data handling efficiently,the hybrid-offload model can achieve the best performance and scalability.These different programming models can be applied and extended to other types of geocomputation to handle big earth data.展开更多
基金Acknowledgements This work was partially supported by the National High-tech R＆D Program of China （863 Program） （2012AA01A301, 2013AA014301, 2013AA01A208）, and by the National Basic Research Program of China （973 Program） （2011CB309705）, and by the National Natural Science Foundation of China （Grant Nos. 61120106005, 61303063 and 61272482）.
文摘Interconnection network plays an important role in scalable high performance computer （HPC） systems. The TH Express-2 interconnect has been used in MilkyWay-2 system to provide high-bandwidth and low-latency interprocessot communications, and continuous efforts are devoted to the development of our proprietary interconnect. This paper describes the state-of-the-art of our proprietary interconnect, especially emphasizing on the design of network interface. Several key features are introduced, such as user-level communication, remote direct memory access, offload collective operation, and hardware reliable end-to-end communication, etc. The design of a low level message passing infrastructures and an upper message passing services are also proposed. The preliminary performance results demonstrate the efficiency of the TH interconnect interface.
基金the National Natural Science Foundation of China under Grant No. U1805262
文摘Due to the rapid development of the maritime networks, there has been a growing demand for computation-intensive applications which have various energy consumption, transmission bandwidth and computing latency requirements. Mobile edge computing(MEC) can efficiently minimize computational latency by offloading computation tasks by the terrestrial access network. In this work, we introduce a space-air-ground-sea integrated network architecture with edge and cloud computing components to provide flexible hybrid computing service for maritime service. In the integrated network, satellites and unmanned aerial vehicles(UAVs) provide the users with edge computing services and network access. Based on the architecture, the joint communication and computation resource allocation problem is modelled as a complex decision process, and a deep reinforcement learning based solution is designed to solve the complex optimization problem. Finally, numerical results verify that the proposed approach can improve the communication and computing efficiency greatly.
文摘With the increasing demand for data traffic and with the massive foreseen deployment of the Internet of Things (IoT), higher data rates and capacity are required in mobile networks. While Heterogeneous Networks (HetNets) are under study toward 5G technology, Wireless Fidelity (WiFi) Access Points (APs) are considered a potential layer within those multiple Radio Access Technologies (RATs). For this purpose, we have proposed in this paper a novel WiFi dimensioning method, to offload data traffic from Long Term Evolution (LTE) to WiFi, by transferring the LTE energy consuming heavy users, to the WiFi network. First, we have calculated the remaining available capacity of the WiFi network based on the estimated load of each WiFi physical channel using the overlapping characteristic of the channels. Then, we were able through this dimensioning method, to calculate the minimum needed number of WiFi APs that ensure the same or better throughput for the LTE transferred users. By this method, we have ensured additional capacity in the LTE network with minimum investment cost in the WiFi network. Finally, we have estimated the profit sharing between LTE and WiFi by considering data bundles subscription revenues and the infrastructure capital and operational costs. We have calculated for each network the profit share using a coalition game theory Shapley value that pinpoints the benefit of the cooperation using the proposed dimensioning method.
文摘近十年来,随着网络技术的发展,网络带宽迅速增长,而同期CPU的性能未得到相应的提高。在吉比特网络下,网络终端CPU处理TCP/IP协议的能力已经成为限制网络应用的瓶颈。为了使终端用户能充分利用广阔的带宽资源,需要提高网络终端的协议处理能力。文中基于FPGA的硬件设计,将原来由软件完成的IP层协议功能完全卸载出来,向CPU提供硬件支持。并且通过功能仿真、综合后仿真、布局布线后仿真验证了设计的可行性,由静态时序分析可知,协议处理器的时钟频率可达50 MH z,处理IP数据流的能力可以达到1.6 G b/s的网络线速度。
基金This research was partially supported by the National Science Foundation through the award SMA-1416509.
文摘Digital earth science data originated from sensors aboard satellites and platforms such as airplane,UAV,and mobile systems are increasingly available with high spectral,spatial,vertical,and temporal resolution data.When such big earth science data are processed and analyzed via geocomputation solutions,or utilized in geospatial simulation or modeling,considerable computing power and resources are necessary to complete the tasks.While classic computer clusters equipped by central processing units(CPUs)and the new computing resources of graphics processing units(GPUs)have been deployed in handling big earth data,coprocessors based on the Intel’s Many Integrated Core(MIC)Architecture are emerging and adopted in many high-performance computer clusters.This paper introduces how to efficiently utilize Intel’s Xeon Phi multicore processors and MIC coprocessors for scalable geocomputation and geo-simulation by implementing two algorithms,Maximum Likelihood Classification(MLC)and Cellular Automata(CA),on supercomputer Beacon,a cluster of MICs.Four different programming models are examined,including(1)the native model,(2)the offload model,(3)the symmetric model,and(4)the hybrid-offload model.It can be concluded that while different kinds of parallel programming models can enable big data handling efficiently,the hybrid-offload model can achieve the best performance and scalability.These different programming models can be applied and extended to other types of geocomputation to handle big earth data.
