专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

71. 发明申请

US20120039161A1 AUTOMATED TRAFFIC ENGINEERING FOR FAT TREE NETWORKS 有权
标题翻译：用于FAT树网络的自动交通工程
公开(公告)号：US20120039161A1
公开(公告)日：2012-02-16
申请号：US12983009
申请日：2010-12-31
申请人： DAVID I. ALLAN , SCOTT A. MANSFIELD
发明人： DAVID I. ALLAN , SCOTT A. MANSFIELD
IPC分类号： H04L12/26 , H04L12/28
CPC分类号： H04L45/125 , H04L45/06 , H04L45/22 , H04L45/28 , H04L45/48 , H04L47/125
摘要： Embodiments of a method implemented in at least one fat tree network node for improved load distribution, wherein the node is one of a plurality of fat tree network nodes in a fat tree network each of which implement a tie-breaking process to produce minimum cost trees, is described. In some embodiments, a spanning tree computation for each root node of the fat tree network in order from a lowest ranked root node to a highest ranked node is performed, a filtering database for each root node of the fat tree network, wherein the filtering database includes a set of media access control (MAC) addresses of the leaf nodes of the fat tree network generated, and link utilization for each computed tree to use as a prefix to link identifiers used for at least one tie-breaking algorithm added.
摘要翻译：在至少一个胖树网络节点中实现的用于改善负载分布的方法的实施例，其中所述节点是胖树网络中的多个胖树网络节点之一，每个胖树网络节点实施断开过程以产生最小成本树，被描述。在一些实施例中，执行从最低排名根节点到最高排名节点的胖树网络的每个根节点的生成树计算，胖树网络的每个根节点的过滤数据库，其中过滤数据库包括生成的胖树网络的叶节点的一组媒体访问控制（MAC）地址，以及用于每个计算树的链路利用率以用作链接用于添加的至少一个断开算法的标识符的前缀。

72. 发明申请

US20110149981A1 DEADLOCK PREVENTION IN DIRECT NETWORKS OF ARBITRARY TOPOLOGY 有权
标题翻译：直接网络中的死亡预防措施
公开(公告)号：US20110149981A1
公开(公告)日：2011-06-23
申请号：US12643280
申请日：2009-12-21
申请人： Peter Michael Klausler
发明人： Peter Michael Klausler
IPC分类号： H04L12/56
CPC分类号： H04L47/12 , G06F15/17312 , G06F15/17381 , H04L45/06 , H04L45/18 , H04L45/60 , H04L49/253
摘要： Aspects of the invention pertain to routing packets in a computer system while avoiding deadlock. A turn rule is set according to unique identifiers associated with switches in the system. Numeric values of switches in possible turns are compared to determine whether a turn is permissible. The rule applies to all nodes in the system. The rule may be violated when using virtual channels. Here, a violation is permissible when using monotonically increasing virtual channel numbers or monotonically decreasing virtual channel numbers. Alternatively, the violations of the turn rule may be allowed if they force a packet to change to a later virtual channel in some fixed ordering of virtual channels. Deadlock can thus be avoided in many different types of architectures, including mesh, torus, butterfly and flattened butterfly configurations.
摘要翻译：本发明的方面涉及在计算机系统中路由分组，同时避免死锁。根据与系统中的开关相关联的唯一标识符设置转弯规则。比较可能转弯中开关的数值，以确定转弯是否允许。该规则适用于系统中的所有节点。使用虚拟通道时可能会违反规则。这里，当使用单调递增的虚拟通道号或单调递减虚拟通道号时，允许违规。或者，如果在某些固定的虚拟通道排序中强制分组改变为稍后的虚拟通道，则可能允许违反转弯规则。因此，可以在许多不同类型的架构中避免死锁，包括网格，环面，蝴蝶和扁平蝶形配置。

73. 发明授权

US07957400B2 Hierarchical network topology 失效
标题翻译：分层网络拓扑
公开(公告)号：US07957400B2
公开(公告)日：2011-06-07
申请号：US12412253
申请日：2009-03-26
申请人： Ronald Michael Henry , Veronica Frances Merryfield
发明人： Ronald Michael Henry , Veronica Frances Merryfield
IPC分类号： H04L12/28
CPC分类号： H04L12/462 , H04L29/12207 , H04L29/12801 , H04L45/04 , H04L45/06 , H04L61/20 , H04L61/6004
摘要： A network including a hierarchical structure of nodes is described. The structure of nodes includes n layers including n−1 layers of switch nodes and 1 layer of computational nodes. Each layer in the structure can include mn-L nodes grouped into units, where m represents a number of nodes in a unit and L represents a layer in the structure, where L=0 represents a lowest layer and L=n−1 represents a highest layer. Each node in a layer other than the computational layer can include a switch node for a unit in a next lower layer. For each unit, each node can be connected to each other node by a point to point link and to a local switch node for the unit by a point to point link. Each node can be connected to each other node and to the local switch node by a local broadcast network for the unit.
摘要翻译：描述了包括节点的分层结构的网络。节点结构包括n层，包括n层交换节点和1层计算节点。结构中的每层可以包括以单位分组的mn-L个节点，其中m表示单元中的节点数，L表示结构中的层，其中L = 0表示最低层，L = n-1表示最高层。除了计算层之外的层中的每个节点可以包括用于下一较低层中的单元的交换节点。对于每个单元，每个节点可以通过点对点链路连接到彼此节点，并且可以通过点对点链路连接到单元的本地交换节点。每个节点可以通过该单元的本地广播网络连接到彼此的节点和本地交换机节点。

74. 发明申请

US20100246437A1 Network Topology 失效
标题翻译：网络拓扑
公开(公告)号：US20100246437A1
公开(公告)日：2010-09-30
申请号：US12412253
申请日：2009-03-26
申请人： Ronald Michael Henry , Veronica Frances Merryfield
发明人： Ronald Michael Henry , Veronica Frances Merryfield
IPC分类号： H04L12/28
CPC分类号： H04L12/462 , H04L29/12207 , H04L29/12801 , H04L45/04 , H04L45/06 , H04L61/20 , H04L61/6004
摘要： A network including a hierarchical structure of nodes is described. The structure of nodes includes n layers including n−1 layers of switch nodes and 1 layer of computational nodes. Each layer in the structure can include mn-L nodes grouped into units, where m represents a number of nodes in a unit and L represents a layer in the structure, where L=0 represents a lowest layer and L=n−1 represents a highest layer. Each node in a layer other than the computational layer can include a switch node for a unit in a next lower layer. For each unit, each node can be connected to each other node by a point to point link and to a local switch node for the unit by a point to point link. Each node can be connected to each other node and to the local switch node by a local broadcast network for the unit.
摘要翻译：描述了包括节点的分层结构的网络。节点结构包括n层，包括n层交换节点和1层计算节点。结构中的每层可以包括以单位分组的mn-L个节点，其中m表示单元中的节点数，L表示结构中的层，其中L = 0表示最低层，L = n-1表示最高层。除了计算层之外的层中的每个节点可以包括用于下一较低层中的单元的交换节点。对于每个单元，每个节点可以通过点对点链路连接到彼此节点，并且可以通过点对点链路连接到单元的本地交换节点。每个节点可以通过该单元的本地广播网络连接到彼此的节点和本地交换机节点。

75. 发明申请

US20090323544A1 INTERNET ROUTE DEAGGREGATION AND ROUTE SELECTION PREFERENCING 有权
标题翻译：互联网路由选择和路由选择优先
公开(公告)号：US20090323544A1
公开(公告)日：2009-12-31
申请号：US12489000
申请日：2009-06-22
申请人： Michael E. Gaddis , Peter N. Hicks , David Barmann , Steven T. Nunes
发明人： Michael E. Gaddis , Peter N. Hicks , David Barmann , Steven T. Nunes
IPC分类号： H04L12/26
CPC分类号： H04L45/06 , H04L45/02 , Y10S707/99931
摘要： A method and system for managing the routing of traffic within a network develops a topological address space map of the network to enable a “best route” selection process. The network is comprised of a backbone connected to a plurality of peering partners. Points on the network monitor traffic flows. A central facility analyzes the traffic flows and routes within the network and performs intelligent routing management.Intelligent routing management ensures that traffic is properly routed through preferred routes on the network, and avoids inefficient routing. Intelligent routing management also selects new routes to be injected into the network in order to further improve the accuracy of the address space map of the network. Intelligent routing management ensures that bandwidth is requested and delivered topologically closely to peering partner networks, and that traffic is carried by the backbone for long haul data distribution in both directions.
摘要翻译：用于管理网络中的业务的路由的方法和系统开发网络的拓扑地址空间映射以启用“最佳路由”选择过程。网络由连接到多个对等伙伴的骨干组成。网络上的点监控流量。中央设施分析网络内的流量和路由，进行智能路由管理。智能路由管理确保通过网络上的首选路由正确路由流量，避免路由效率低下。智能路由管理还选择要注入网络的新路由，以进一步提高网络地址空间映射的准确性。智能路由管理确保带宽被请求和拓扑结构紧密地传递给对等伙伴网络，并且流量由主干承载，用于沿双向的长途数据分发。

76. 发明授权

US07587516B2 Class network routing 失效
标题翻译：类网络路由
公开(公告)号：US07587516B2
公开(公告)日：2009-09-08
申请号：US10468999
申请日：2002-02-25
申请人： Gyan Bhanot , Matthias A. Blumrich , Dong Chen , Paul W. Coteus , Alan G. Gara , Mark E. Giampapa , Philip Heidelberger , Burkhard D. Steinmacher-Burow , Todd E. Takken , Pavlos M. Vranas
发明人： Gyan Bhanot , Matthias A. Blumrich , Dong Chen , Paul W. Coteus , Alan G. Gara , Mark E. Giampapa , Philip Heidelberger , Burkhard D. Steinmacher-Burow , Todd E. Takken , Pavlos M. Vranas
IPC分类号： G06F15/173 , H04L12/66 , H04L12/50
CPC分类号： H04L45/16 , H04L45/06
摘要： Class network routing is implemented in a network such as a computer network comprising a plurality of parallel compute processors at nodes thereof. Class network routing allows a compute processor to broadcast a message to a range (one or more) of other compute processors in the computer network, such as processors in a column or a row. Normally this type of operation requires a separate message to be sent to each processor. With class network routing pursuant to the invention, a single message is sufficient, which generally reduces the total number of messages in the network as well as the latency to do a broadcast. Class network routing is also applied to dense matrix inversion algorithms on distributed memory parallel supercomputers with hardware class function (multicast) capability. This is achieved by exploiting the fact that the communication patterns of dense matrix inversion can be served by hardware class functions, which results in faster execution times.
摘要翻译：在诸如包括在其节点处的多个并行计算处理器的计算机网络的网络中实现类网络路由。类网络路由允许计算处理器将消息广播到计算机网络中的其他计算处理器的范围（一个或多个），例如列或行中的处理器。通常这种类型的操作需要单独的消息发送到每个处理器。根据本发明的类网络路由，单个消息是足够的，这通常减少了网络中的消息总数以及进行广播的延迟。类网络路由也适用于具有硬件类功能（组播）能力的分布式存储并行超级计算机上的密集矩阵求逆算法。这是通过利用密集矩阵反演的通信模式可以通过硬件类功能来实现的，这导致更快的执行时间。

77. 发明申请

US20090201925A1 Apparatus and method for setting routing path between routers in chip 审中-公开
公开(公告)号：US20090201925A1
公开(公告)日：2009-08-13
申请号：US12385622
申请日：2009-04-14
申请人： Sang-Woo Rhim , Beom-Hak Lee , Jae-Kon Lee , Eul-Seok Kim
发明人： Sang-Woo Rhim , Beom-Hak Lee , Jae-Kon Lee , Eul-Seok Kim
IPC分类号： H04L12/56
CPC分类号： G06F15/16 , H04L45/00 , H04L45/06 , H04L49/109 , H04L49/15
摘要： An apparatus and a method for setting a routing path in System-on-a-Chip (SoC) having an n×n net topology-based structure comprising a plurality of intellectual properties (IPs), each with a unique address, and a plurality of switches forming one-to-one correspondence to the IPs, to transmit and receive data between the IPs by using at least one of the switches. Accordingly, an orthogonal code having orthogonality is assigned according to the direction of transmission of each data (that is, output port). Then, an output port where the data is transmitted is determined, and the data is spread based on an orthogonal code assigned to the output ports of the at least one of the switches.

78. 发明申请

US20090198956A1 System and Method for Data Processing Using a Low-Cost Two-Tier Full-Graph Interconnect Architecture 审中-公开
标题翻译：使用低成本两层全图互连架构的数据处理系统和方法
公开(公告)号：US20090198956A1
公开(公告)日：2009-08-06
申请号：US12024662
申请日：2008-02-01
申请人： Lakshminarayana B. Arimilli , Ravi K. Arimilli , Ramakrishnan Rajamony , Edward J. Seminaro , William E. Speight
发明人： Lakshminarayana B. Arimilli , Ravi K. Arimilli , Ramakrishnan Rajamony , Edward J. Seminaro , William E. Speight
IPC分类号： G06F15/80 , G06F15/76 , G06F9/00
CPC分类号： H04L45/12 , H04L45/06 , H04L49/109 , H04L49/25
摘要： A system and method are provided for implementing a two-tier full-graph interconnect architecture. In order to implement a two-tier full-graph interconnect architecture, a plurality of processors are coupled to one another to create a plurality of supernodes. Then, the plurality of supernodes are coupled together to create the two-tier full-graph interconnect architecture. Data is then transmitted from one processor to another within the two-tier full-graph interconnect architecture based on an addressing scheme that specifies at least a supernode and a processor chip identifier associated with a target processor to which the data is to be transmitted.
摘要翻译：提供了一种用于实现两层全图互连架构的系统和方法。为了实现两层全图互连架构，多个处理器彼此耦合以创建多个超节点。然后，多个超节点被耦合在一起以创建双层全图互连体系结构。基于指定至少一个超级节点的寻址方案和与要发送数据的目标处理器相关联的处理器芯片标识符，在两层全图互连架构内将数据从一个处理器发送到另一个处理器。

79. 发明授权

US07554930B2 Internet route deaggregation and route selection preferencing 有权
标题翻译：互联网路由解聚和路由选择优先
公开(公告)号：US07554930B2
公开(公告)日：2009-06-30
申请号：US11084804
申请日：2005-03-18
申请人： Michael E. Gaddis , Peter N. Hicks , David Barmann , Steven T. Nunes
发明人： Michael E. Gaddis , Peter N. Hicks , David Barmann , Steven T. Nunes
IPC分类号： H04L12/28 , H04L12/56
CPC分类号： H04L45/06 , H04L45/02 , Y10S707/99931
摘要： A method and system for managing the routing of traffic within a network develops a topological address space map of the network to enable a “best route” selection process. The network is comprised of a backbone connected to a plurality of peering partners. Points on the network monitor traffic flows. A central facility analyzes the traffic flows and routes within the network and performs intelligent routing management.Intelligent routing management ensures that traffic is properly routed through preferred routes on the network, and avoids inefficient routing. Intelligent routing management also selects new routes to be injected into the network in order to further improve the accuracy of the address space map of the network. Intelligent routing management ensures that bandwidth is requested and delivered topologically closely to peering partner networks, and that traffic is carried by the backbone for long haul data distribution in both directions.
摘要翻译：用于管理网络中的业务的路由的方法和系统开发网络的拓扑地址空间映射以启用“最佳路由”选择过程。网络由连接到多个对等伙伴的骨干组成。网络上的点监控流量。中央设施分析网络内的流量和路由，进行智能路由管理。智能路由管理确保通过网络上的首选路由正确路由流量，避免路由效率低下。智能路由管理还选择要注入网络的新路由，以进一步提高网络地址空间映射的准确性。智能路由管理确保带宽被请求和拓扑结构紧密地传递给对等伙伴网络，并且流量由主干承载，用于沿双向的长途数据分发。

80. 发明申请

US20090067445A1 ROUTER AND ROUTING NETWORK 审中-公开
标题翻译：路由器和路由网络
公开(公告)号：US20090067445A1
公开(公告)日：2009-03-12
申请号：US12091830
申请日：2006-10-27
申请人： Jean-Philippe Diguet , Samuel Evain
发明人： Jean-Philippe Diguet , Samuel Evain
IPC分类号： H04L12/56
CPC分类号： H04L45/06
摘要： A router in which each of the input/output ports is associated with identifiers enabling each of the other input/output ports to locate it according to a code specific to each of them, comprising means for identifying in an incident packet a routing instruction indicating a forward identifier of the desired output port; and means for sending back the packet where the forward identifier is suppressed from the routing instruction and a return identifier is inserted therein.
摘要翻译：一种路由器，其中每个输入/输出端口与标识符相关联，使得每个其他输入/输出端口能够根据其中每个输入/输出端口的特定代码来定位它，包括用于在事件包中识别指示所需输出端口的正向标识符; 以及用于从所述路由指令发回所述前向标识符被抑制的分组并且在其中插入返回标识符的装置。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式