The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
The following topics are dealt with: interconnects and switching; input/output and networking; data center bridging;interconnects for exascale computing; and silicon photonics.
Summary form only given. The industry is undergoing a disruption to construct modern data centers and cloud designs. Denser server and storage, place demands on the networks to keep up with the bottlenecks. CPUs are faster, projecting to achieve 5 Ghz and CPUs are wider-with more cores. The advent of virtualization consolidates servers across workloads but also creates new challenges with network...
Single processor performance has exhibited substantial growth over the last three decades [1] as shown in Figure 1. What is also desired are techniques which enable connecting together multiple processors in order to create scalable, modular and resilient multiprocessor systems. Beginning with the production of the Intel® Xeon® processor 5500 series, (previously codenamed “Nehalem-EP”), the Intel®...
We consider the problem of scheduling packets transmission across the switching fabric in an input-queued switch. In order to achieve maximum throughput, scheduling algorithms usually employ the queue length as a parameter for determining the priority to serve a given queue. Unfortunately, optimal schedulers are too complex to be implemented directly in hardware, which is the only viable solution...
Fat-tree networks are the most popular topology among indirect networks in today's supercomputers. Current supercomputers are generally operated in a shared environment under the control of a job scheduler, executing many parallel applications simultaneously. The competition between these applications to use the same network resources causes a degradation in the applications' performance. The application...
We have achieved sharing a single-root I/O virtualization (SR-IOV) compliant PCI Express (PCIe) I/O device among multiple computers. A device share not only inside a single computer among virtual machines, but also among multiple computers attracts a great interest because it provides efficient utilization of computer resources. Because PCIe is originally a single-root system, realizing multi-root...
With the rapid evolution of network speed from 1Gbps to 10Gbps, a wide spectrum of research has been done on TCP/IP to improve its processing efficiency on general purpose processors. However, most of them did studies only from the performance perspective and ignored its power efficiency. As power has become a major concern in data centers, where servers are often interconnected with 10GbE, it becomes...
Collective communication operations provided by The Message Passing Interface (MPI) are heavily used by scientific applications at large scale. The current MPI standard, MPI-2.2, only defines blocking collective communication calls, which does not allow simultaneous computation and communication. It is expected that MPI-3 will allow for non-blocking collective communication. The newly introduced ConnectX-2...
In order to fully realize the potential of Cloud and High Performance Computing (HPC) applications, significant improvement is required in the cost/performance of data center networks. While recent industry standards such as Quantized Congestion Notification (QCN) for Converged Enhanced Ethernet (CEE) have begun to address this issue, there are still significant problems left open. Therefore we propose...
Data Center Networks represent the convergence of computing and networking, of data and storage networks, and of packet transport mechanisms in Layers 2 and 3. Congestion control algorithms are a key component of data transport in this type of network. Recently, a Layer 2 congestion management algorithm, called QCN (Quantized Congestion Notification), has been adopted for the IEEE 802.1 Data Center...
Data center networks (DCNs) usually have a regular topology providing high bandwidth and low latency with multi-paths. In a data-intensive application on DCNs, it is important to be able to use all the available bandwidth. However, this is not easy to achieve because of the locality of congestion information. In this paper, we propose a reactive reroute strategy to incorporate with IEEE 802.1Qau standards...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.