Continuing to integrate more aggregate bandwidth and higher radix into switch devices is an economic imperative because it creates value both for the supplier and customer in large data center environments which are an increasingly important part of the marketplace. While new silicon processes continue to shrink transistor and other chip feature dimensions, process technology cannot be relied upon as a key driver of power reduction. Transitioning from 28 nm to 16 nm is a special case where FinFET provides additional power scaling, but subsequent FinFET nodes are not expected to deliver as substantial of power reductions to meet the desired increases in integration. The disclosed switch architecture attacks the power consumption problem by controlling the rate at which power-consuming activities occur.

A switching device includes a processor, an input buffer, an output buffer, and a Banyan switching architecture, where the processor is configured to convert an initial switching table to a non-congestion switching table and an order-adjustment table using a preset reordering algorithm; the input buffer is configured to save first period data that is from an input port; the processor is further configured to perform, using the non-congestion switching table, data switching on data in the first full-period data saved in the input buffer, to obtain second full-period data; the Banyan switching architecture is configured to perform synchronous data switching on the second full-period data; the output buffer is configured to save the second full-period data on which the synchronous data switching has been performed; the processor is further configured to adjust, using the order-adjustment table, a data order of the second period data.

A programmable switch includes a plurality of ports for communicating with devices on a network. Circuitry of the programmable switch is configured to receive a series of related messages from a first device on the network via at least one port, and determine whether one or more messages of the series of related messages have been received out-of-order based at least in part on a sequence number included in the one or more messages. The series of related messages are sent by the programmable switch to a second device via one or more ports in an order indicated by sequence numbers included in the series of related messages by delaying at least one message. According to one aspect, a network controller selects a programmable switch between the first device and the second device to serve as a message sequencer for reordering out-of-order messages using a stored network topology.

A data transmission method, including obtaining by a transmit end, at least two to-be-transmitted packets, determining a first interface used to transmit each of the packets in at least two interfaces of the transmit end, and determining an identifier of each of the packets that is related to the first interface, where the identifier represents an order of the first interface used to transmit each of the packets in the at least two interfaces used to send the at least two packets adding the identifier to a packet header of each of the packets and sending a packet added with the identifier to the receive end device through the first interface, so that the receive end device adjusts, based on the identifier, an order of the packet added with the identifier.

Methods and systems for managing multiple transmit queues of a networking device of a host machine in a virtual machine system. The networking device includes multiple transmit queues that are used by multiple guests of the virtual machine system for the transmission of packets in a data communication. A hypervisor of the virtual machine system manages the switching from one or more transmit queues (i.e., old transmit queues) to one or more other queues (i.e., new transmit queues) by managing a flow of packets in the virtual machine system to maintain a proper sequence of packets and avoid a need to re-order the transmitted packets at a destination.

Disclosed are an apparatus and method for reconstructing a transmitted file with high performance in real time, which select analysis target packets for reconstruction by first checking using hardware whether data file-related information is present in packets transmitted via large-capacity traffic over a broadband network, and which reconstruct a file in real time only from the selected analysis target packets. The file reconstruction apparatus for reconstructing a data file from packets on a network includes a packet monitoring unit for extracting packets on the network, a collected packet selection unit for determining whether, for the extracted packets, each packet is a reconstruction target based on flow information, and selecting a reconstruction target packet, and a file reconstruction unit for performing file reconstruction by extracting data from the reconstruction target packet and by storing the extracted data as data of a reconstructed file in a relevant flow.

Deterministic real-time multi-protocol heterogeneous packet-based transport is achieved by traffic shaping. When receiving a plurality of packets from a root complex where contents of each packet from the plurality of packets organized in accordance with a first protocol, a sequence number is added to each packet and a packet type is identified. Every packet in the first plurality of packets is encapsulated into at least one packet organized in accordance with a second protocol to form a second plurality of packets organized in accordance with the second protocol. All the packets from the second plurality of packets pass traffic scheduling or traffic shaping prior being sent via a plurality of connections to avoid burstiness and to achieve bounded transport latency in the plurality of connections, thereby providing deterministic real-time behavior in distributed systems.

Deterministic real-time multi-protocol heterogeneous packet-based transport is achieved by traffic shaping. When receiving a plurality of packets from a root complex where contents of each packet from the plurality of packets organized in accordance with a first protocol, a sequence number is added to each packet and a packet type is identified. Every packet in the first plurality of packets is encapsulated into at least one packet organized in accordance with a second protocol to form a second plurality of packets organized in accordance with the second protocol. All the packets from the second plurality of packets pass traffic scheduling or traffic shaping prior being sent via a plurality of connections to avoid burstiness and to achieve bounded transport latency in the plurality of connections, thereby providing deterministic real-time behavior in distributed systems.

A method of cloning and mangling a received data packet in which an unused space of a receiving buffer can be used to accommodate at least some generated clone packets. Additional memory-use efficiencies can be realized by employing scatter-gather lists in the process of clone-packet generation when the size of the received data packet exceeds a predetermined threshold size. The method enables the corresponding network device to improve the packet-processing speed and memory use compared to those achievable with the use of conventional methods.

Apparatuses, methods, and systems for dynamically controlling a local buffer of a modem of a wireless device are disclosed. One method includes receiving and queuing transmission packets in the local buffer of the modem of the wireless device for wireless transmission to a receiving device, purging each transmission packet from the local buffer after receiving an acknowledgement of reception of the transmission packet from the receiving device, and requesting acknowledgement from the receiving device when a queue of the transmission packets within the local buffer exceeds a threshold level, wherein the receiving device aggregates acknowledgment responses to a plurality of unpurged transmission packets in the local buffer and transmits an aggregated acknowledgment to the modem.