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.
In this work, we seek to gain an understanding of the InfiniBand network processing limitations that might exist in gathering performance metric information from InfiniBand switches using our new LDMS ibfabric sampler. The limitations studied consist of delays in gathering InfiniBand metric information from a specific switch device due to the switch's processor response delays or RDMA contention for...
A trend has emerged toward the development of critical systems hosted on the Ethernet databus. Ethernet enjoys unprecedented standardization in commercial networks, and a vast network of engineers with training and experience. It is now commonly known that implementations of deterministic Ethernet for avionics networks known as ARINC 664 part 7 exist on several commercial airframes. In other platforms,...
Software-defined networking (SDN) is a promising future network architecture which introduces new dimensionsin flexibility and adaptability to cope with different Quality of service (QoS) metrics, such as latency constraint, loss rate, etc. In this paper, we provide differentiated Services for flows of a topic-oriented publish/subscribe system to address some of the QoS guarantee issues, by leveraging...
We tackle the problem of a network switch enforcing fair bandwidth sharing of the same link among many TCP-like senders. Most of the mechanisms to solve this problem are based on complex scheduling algorithms, whose feasibility becomes very expensive with today's line rate requirements, i.e. 10–100 Gbit/s per port. We propose a new scheme called FDPA in which we do not modify the scheduler, but instead...
Among the low-diameter, high-radix networks beingdeployed in next-generation HPC systems, dual-rail fat-treenetworks are a promising approach. Adding additional injectionconnections (rails) to one or more network planes allows multirailfat-tree networks to alleviate communication bottlenecks. These multi-rail networks necessitate new design considerations, such as routing choices, job placements,...
High-radix routers with low latency and high bandwidth play an increasingly important role in the design of large-scale interconnection networks such as those used in super-computers and datacenters. The tile-based crossbar approach partitions a single large crossbar into many small tiles and can considerably reduce the complexity of arbitration while providing throughput higher than the conventional...
Software-Defined Networking (SDN) provides an abstraction of the network that allows network administrators to configure network devices seamlessly across the enterprise networks. Although SDN has created a new paradigm in next generation network, centralized controller architecture in SDN imposes overhead issue because control plane will handle all flows regardless of size and priority. Besides that,...
Recent research has shown that the integration of a custom-silicon network-on-chip (NoC) into an FPGA fabric can significantly help on-chip communication bandwidth. Although not appropriate for all communication, FPGA hard NoCs provide a scalable infrastructure that will increase in importance as FPGA sizes increase. To date, most FPGA hard NoC implementation has focused on packet-switched routing...
We focus on designing efficient integrated schedulers for handling mixed unicast and multicast traffic. We consider an input-queued switch with a multicast-capable switch fabric. At each input port of the switch, there are N dedicated unicast VOQs and one shared multicast queue (MQ). An existing approach to the design of integrated scheduler (i.e., a sequential scheduler) is to run two component schedulers,...
Virtual machine (VM) consolidation and switch/path consolidation are two typical techniques for improving energy efficiency in data centers (DCs). Most of existing work separately optimize VM consolidation and switch consolidation which results in inferiority of the optimization performance. Moreover, these work usually handle a user application as a VM flow (i.e., a source VM is connected to a destination...
This paper presents a comparison of throughputand end-to-end latency of synchronous and asynchronous heterogeneousNoC under uniform and exponential traffic conditionsusing different parameters. The parameters we have chosen areno. of cores, load (traffic) and no. of VCs of a router. Further, sinkbandwidth analysis of synchronous and asynchronous NoC underuniform traffic was studied and compared. The...
MeteoSwiss, the Swiss national weather forecast institute, has selected densely populated accelerator servers as their primary system to compute weather forecast simulation. Servers with multiple accelerator devices that are primarily connected by a PCI-Express (PCIe) network achieve a significantly higher energy efficiency. Memory transfers between accelerators in such a system are subjected to PCIe...
SpaceFibre [1][2][3] is the next generation of SpaceWire [4] on-board data-handling network technology for spaceflight operations, which runs over both electrical and fibre optic media. SpaceFibre has many benefits compared to SpaceWire, including much higher data-rates, integrated quality of service, fault recovery capabilities, multi-laning with graceful degradation and hot and cold redundancy,...
Current network protocols are exceedingly strict, making the administrative duties of network providers difficult. The switches that control a network rely on stringent definitions, usually defined by the company who designed the switch, with which they can forward packets. Software Defined Networking (SDN) is a concept which allows control over the operations of a switch via remote software applications...
Network-on-Chip is an emerging paradigm for integrating very high number of Intellectual Property blocks on a single Integrated Chip. In this paper, crossbar switch is designed for 2-D Mesh Network-on-Chip to meet the current requirements of high speed networks. The design reduced the power consumption significantly and can meet the scalable bandwidth. The functional verification of the crossbar switch...
The growth of bandwidth in computer networks fuels the constant adoption of measurement equipment and methodologies. Networking equipment offers processing capacity in the Terabit/s range nowadays. However, test equipment in academia lags behind. We propose FLOWer, a novel and cost effective approach capable of testing such high-speed devices. FLOWer combines an inexpensive software packet generator...
To reduce the energy consumption of a large number of network devices in a data center, energy-efficient schemes use various heuristics to consolidate traffic to fewer switches. Most of these works, however, ignore the flow-level performance, which is one of the most critical requirements in production data centers. Hence, flow rate allocation should be considered together with flow path selection...
For years, Virtual Network Operators (VNOs) are a major player in the satellite communication market landscape. Typically, they repackage services leased from Satellite Network Operators (SNOs) to provide their customers with added-value end-to-end services that mostly go beyond the satellite network and cover terrestrial networks. However, the level of control and visibility that VNOs have on their...
An effort to deploy security devices by the network provider has been increasing as the network is being exposed to various types of network attacks. However, network providers are incapable of handling all types of attacks as each security device is designed for a certain purpose. If an attack breaks out, only one particular device becomes busy in terms of resource usage while others being idle....
3D integration is a practical solution for overcoming the failure of Dennard scaling in future technology generations. This emerging technology stacks several die slices on top of each other on a single chip in order to provide higher-bandwidth and lower-latency than a 2D design due to extremely shorter inter-layer distances in the third dimension and. In this paper, we leverage the low-latency vertical...
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.