Method for maintaining cache coherence using a distributed directory with event driven updates
    1.
    发明授权
    Method for maintaining cache coherence using a distributed directory with event driven updates 失效
    使用事件驱动更新的分布式目录来维护高速缓存一致性的方法

    公开(公告)号:US08195890B1

    公开(公告)日:2012-06-05

    申请号:US11895052

    申请日:2007-08-22

    IPC分类号: G06F12/00

    CPC分类号: G06F12/0824

    摘要: The present invention is a protocol for maintaining cache consistency between multiprocessors within a tightly coupled system. A distributed directory is maintained within the data-sharing processors, so that copies can be invalidated when modified. All transfers are event driven, rather the polled, to reduce bus-bandwidth consumption. Deadlocks are avoided by placing to-be-executed command codes in the returned response packets, when the request-forwarding queues are full or not present.

    摘要翻译: 本发明是用于在紧耦合系统内的多处理器之间保持高速缓存一致性的协议。 在数据共享处理器中维护一个分布式目录,以便在修改时可以使副本无效。 所有传输都是事件驱动的,而不是轮询,以减少总线带宽消耗。 当请求转发队列已满或不存在时,通过在返回的响应数据包中放置要执行的命令代码来避免死锁。

    METHOD FOR PACING MULTI-RATE TIME SENSITIVE TRAFFIC ON ETHERNET AND BRIDGE CONNECTED NETWORKS
    2.
    发明申请
    METHOD FOR PACING MULTI-RATE TIME SENSITIVE TRAFFIC ON ETHERNET AND BRIDGE CONNECTED NETWORKS 失效
    用于在以太网和桥连接网络上进行多速时间敏感性交通的方法

    公开(公告)号:US20110116371A1

    公开(公告)日:2011-05-19

    申请号:US13014132

    申请日:2011-01-26

    IPC分类号: H04L12/26

    摘要: A system and method for delaying packet transmissions within source devices and network bridges for the purpose of reducing the worst-case delays associated with forwarding of time-sensitive packets through a bridged network. The minimal bridged network system include a first device node attached to a first network link, a second device node attached to a second network link, and a bridge connecting the first network link to the second network link. The method and system includes shaping the traffic within a source device or bridge (shaping involves spreading bunched packets over time). Within a bridge, shaping occurs at each output port of the bridge, applying a distinct shaper to the traffic coming from each of the input ports, as well as distinct shapers for the distinct traffic class coming each of the input ports. According to the system and method disclosed herein, the present invention ensures small latencies when passing through network bridges, regardless of the network topology or other concurrent data-packet transfers, subject to the restriction that the cumulative time-sensitive bandwidth flowing through any link is less than the prenegotiated rate below that of the link's cumulative bandwidth capacity.

    摘要翻译: 用于延迟源设备和网桥内的分组传输的系统和方法,用于减少通过桥接网络转发时间敏感分组的最坏情况延迟。 最小桥接网络系统包括附接到第一网络链路的第一设备节点,附接到第二网络链路的第二设备节点以及将第一网络链路连接到第二网络链路的桥接器。 该方法和系统包括对源设备或网桥内的流量进行整形(整形涉及在一段时间内扩展捆绑包)。 在桥内,在桥的每个输出端口处进行整形,对来自每个输入端口的流量应用不同的整形器,以及用于来自每个输入端口的不同业务类别的不同的整形器。 根据本文所公开的系统和方法,本发明在通过网桥时保证小的延迟,无论网络拓扑或其他并发数据分组传输如何,受限于流经任何链路的累积时间敏感带宽是 小于预先协商的速率低于链路的累积带宽容量。

    High speed transmission system with clock inclusive balanced coding
    3.
    发明授权
    High speed transmission system with clock inclusive balanced coding 有权
    高速传输系统,带时钟包含平衡编码

    公开(公告)号:US06690309B1

    公开(公告)日:2004-02-10

    申请号:US10024609

    申请日:2001-12-17

    IPC分类号: H03M720

    摘要: A data transmission system (100) may include a transmitting portion (102) and a receiving portion (104). A transmitting portion (102) may include an encoder (106) that may encode data values of n bits into codes of m bits, where n is less than m. Codes may be transmitted with corresponding clock values. The absolute value of the DC component of a code summed with a corresponding clock value can be no more tan one for all code values.

    摘要翻译: 数据传输系统(100)可以包括传输部分(102)和接收部分(104)。 发送部分(102)可以包括编码器(106),其可以将n位的数据值编码为m位的码,其中n小于m。 可以用对应的时钟值发送代码。 与对应的时钟值相加的代码的DC分量的绝对值对于所有代码值可以不再是tan 1。

    Method for pacing multi-rate time sensitive traffic on ethernet and bridge connected networks
    4.
    发明授权
    Method for pacing multi-rate time sensitive traffic on ethernet and bridge connected networks 失效
    在以太网和网桥连接网络上进行多速率时间敏感流量的方法

    公开(公告)号:US07881194B1

    公开(公告)日:2011-02-01

    申请号:US11501984

    申请日:2006-08-10

    IPC分类号: G01R31/08

    摘要: A system and method for delaying packet transmissions within source devices and network bridges for the purpose of reducing the worst-case delays associated with forwarding of time-sensitive packets through a bridged network. The minimal bridged network system include a first device node attached to a first network link, a second device node attached to a second network link, and a bridge connecting the first network link to the second network link. The method and system includes shaping the traffic within a source device or bridge (shaping involves spreading bunched packets over time). Within a bridge, shaping occurs at each output port of the bridge, applying a distinct shaper to the traffic coming from each of the input ports, as well as distinct shapers for the distinct traffic class coming each of the input ports. According to the system and method disclosed herein, the present invention ensures small latencies when passing through network bridges, regardless of the network topology or other concurrent data-packet transfers, subject to the restriction that the cumulative time-sensitive bandwidth flowing through any link is less than the prenegotiated rate below that of the link's cumulative bandwidth capacity.

    摘要翻译: 用于延迟源设备和网桥内的分组传输的系统和方法,用于减少通过桥接网络转发时间敏感分组的最坏情况延迟。 最小桥接网络系统包括附接到第一网络链路的第一设备节点,附接到第二网络链路的第二设备节点以及将第一网络链路连接到第二网络链路的桥接器。 该方法和系统包括对源设备或网桥内的流量进行整形(整形涉及在一段时间内扩展捆绑包)。 在桥内,在桥的每个输出端口处进行整形,对来自每个输入端口的流量应用不同的整形器,以及用于来自每个输入端口的不同业务类别的不同的整形器。 根据本文所公开的系统和方法,本发明在通过网桥时保证小的延迟,无论网络拓扑或其他并发数据分组传输如何,受限于流经任何链路的累积时间敏感带宽是 小于预先协商的速率低于链路的累积带宽容量。

    Method for maintaining cache coherence using a distributed directory with event driven updates
    7.
    发明授权
    Method for maintaining cache coherence using a distributed directory with event driven updates 失效
    使用事件驱动更新的分布式目录来维护高速缓存一致性的方法

    公开(公告)号:US08543771B1

    公开(公告)日:2013-09-24

    申请号:US13488080

    申请日:2012-06-04

    IPC分类号: G06F12/00

    CPC分类号: G06F12/0824

    摘要: The present invention is a protocol for maintaining cache consistency between multiprocessors within a tightly coupled system. A distributed directory is maintained within the data-sharing processors, so that copies can be invalidated when modified. All transfers are event driven, rather the polled, to reduce bus-bandwidth consumption. Deadlocks are avoided by placing to-be-executed command codes in the returned response packets, when the request-forwarding queues are full or not present.

    摘要翻译: 本发明是用于在紧耦合系统内的多处理器之间保持高速缓存一致性的协议。 在数据共享处理器中维护分布式目录,以便在修改时可以使副本无效。 所有传输都是事件驱动的,而不是轮询,以减少总线带宽消耗。 当请求转发队列已满或不存在时,通过在返回的响应数据包中放置要执行的命令代码来避免死锁。

    Method and system for providing a communication connection using stream identifiers
    8.
    发明授权
    Method and system for providing a communication connection using stream identifiers 有权
    使用流标识符提供通信连接的方法和系统

    公开(公告)号:US06631415B1

    公开(公告)日:2003-10-07

    申请号:US09531065

    申请日:2000-03-18

    IPC分类号: G06F1516

    摘要: A method and a system for providing a connection using a stream identifier (“id”) are disclosed. In one embodiment, the system identifies a talker node, which provides an isochronous data stream via a communication media. Once the talker node is identified, an identifier and an sub-identifier of the talker node are identified. After identifying the identifier and sub-identifier, the stream id associated with the isochronous data stream is identified in response to the identifier and sub-identifier. Next, the system establishes at least one connection according to the stream id. When the connection between the second and talker nodes is established, the listener node reads the isochronous data stream.

    摘要翻译: 公开了一种使用流标识符(“id”)提供连接的方法和系统。 在一个实施例中,系统识别通过通信介质提供同步数据流的讲话者节点。 一旦识别了讲话者节点,就识别出讲话者节点的标识符和子标识符。 在识别标识符和子标识符之后,响应于标识符和子标识符识别与同步数据流相关联的流id。 接下来,系统根据流ID建立至少一个连接。 当建立第二和讲话者节点之间的连接时,收听节点读取同步数据流。