One embodiment provides a method, including: allocating, at a node within a local network, availability to a broader network into one or more units; associating, using a processor, one or more tokens with the one or more units; distributing, over a network, the one or more tokens to a requesting node; receiving, over a network, the one or more tokens from the requesting node, where the one or more tokens have been associated with network payload; confirming, using the processor, that the one or more tokens are valid; and in response to the confirming, sending the network payload onto the broader network. Other embodiments are described and claimed.

A first request may be received to write a first set of data to a first storage device of a first storage node. The first storage device may be capable of transferring or receiving data directly to or from a second storage device without transferring the data to or from a host device mapped to the first storage node. It may be determined that a first token clash check does not need to occur for the first request. The first token clash check may include determining whether the first request is requesting to write to one or more addresses that are associated with one or more tokens owned by one or more transactions. The one or more tokens may be a proxy for a set of data within one or more particular address ranges of the first storage device.

Technologies for accelerating non-uniform network input/output accesses include a multi-home network interface controller (NIC) of a network computing device communicatively coupled to a plurality of non-uniform memory access (NUMA) nodes, each of which include an allocated number of processor cores of a physical processor package and an allocated portion of a main memory directly linked to the physical processor package. The multi-home NIC includes a logical switch communicatively coupled to a plurality of logical NICs, each of which is communicatively coupled to a corresponding NUMA node. The multi-home NIC is configured to facilitate the ingress and egress of network packets by determining a logical path for each network packet received at the multi-home NIC based on a relationship between one of the NUMA nodes and/or a logical NIC (e.g., to forward the network packet from the multi-home NIC) coupled to the one of the NUMA nodes. Other embodiments are described herein.

Systems, apparatuses, and methods for implementing an asynchronous router with virtual channel (VC) control. The asynchronous router may support multiple VCs for connections to other routers. The asynchronous router may include an interface unit on each switch boundary, with each interface unit including a data merge unit. The data merge unit may include a full detector unit for each VC, with the full detector unit counting the number of flits sent out on a respective VC and counting the number of credits released by the successor router. Whenever the successor router has no credits available, the full detector unit will assert the full signal to prevent any input requests from requesting to transmit over that particular VC. When the full signal is asserted, a timer unit may be activated to repeatedly check if any credits have been released in the successor router.

A crossbar switch has N input ports, M output ports, and a switching matrix with N×M crosspoints. In an embodiment, each crosspoint contains an internal queue (XQ), which can store one or more packets to be routed. Traffic rates to be realized between all Input/Output (IO) pairs of the switch are specified in an N×M traffic rate matrix, where each element equals a number of requested cell transmission opportunities between each IO pair within a scheduling frame of F time-slots. An efficient algorithm for scheduling N traffic flows with traffic rates based upon a recursive and fair decomposition of a traffic rate vector with N elements, is proposed. To reduce memory requirements a shared row queue (SRQ) may be embedded in each row of the switching matrix, allowing the size of all the XQs to be reduced. To further reduce memory requirements, a shared column queue may be used in place of the XQs. The proposed buffered crossbar switches with shared row and column queues, in conjunction with the row scheduling algorithm and the DCS column scheduling algorithm, can achieve high throughput with reduced buffer and VLSI area requirements, while providing probabilistic guarantees on rate, delay and jitter for scheduled traffic flows.

There is provided a transmission apparatus including: a buffer provided with a queue for each of user flows, the buffer configured to store data in the queue for the user flow to which the data belongs; a storage unit configured to store a link list indicating an output order of the user flows; an issuing unit configured to issue a predetermined output permissible amount to the queue for the user flow by referring to the link list; and an output control unit configured to output data from the queue to which the output permissible amount is issued.

A method for managing port bandwidth in network devices. The method includes determining a first and a second ingress bandwidth of a first and a second network chip, respectively, determining an egress bandwidth of an egress port of a third network chip, determining a first and a second weight for the first and the second network chip, respectively, where the first and the second weight are determined based on a bandwidth including the first and second ingress bandwidth, processing a first data packet, received by a first ingress port administrated by the first network chip, based on the first weight and the egress bandwidth, and processing a second data packet, received by a second ingress port administrated by the second network chip, based on the second weight, and the egress bandwidth, where the destination of the first and the second data packet is the egress port.

A rate adaptation system includes a barrel shift slot register and a rate adaptation register. The barrel shift slot register includes a plurality of slots with one of a valid read request or a dummy read request. A rate adaptation register is configured to sequentially cycle through the slots of the barrel shift register in response to a clock providing valid read requests to a FIFO buffer and to skip provision of valid read requests for clock cycles of the first clock associated with slots that include dummy read requests. The rate adaption register may also receive data blocks from the FIFO buffer and provide those data blocks to another FIFO buffer.

A communication device, includes: a plurality of queues each configured to accumulate a packet; a scheduler configured to provide a permissible readout amount to each of the plurality of queues in accordance with an order that is based on a priority of each queue; a read processor configured to read out the packet from the plurality of queues, the permissible readout amount being consumed according to amount of the packets read out; and an accumulation amount counter configured to count an accumulation amount of the packets accumulated in each of the plurality of queues, wherein the accumulation amount counter notifies the scheduler of a change in the accumulation amount, and wherein the scheduler adjusts the priority of, among the plurality of queues, the queue of which the accumulation amount has changed, in response to the notification from the accumulation amount counter.

Systems, methods, and computer programs are presented for managing network traffic. A network switch includes a switch fabric and a resource coherency and analytics engine (RCAE) coupled to the switch fabric. The RCAE includes one or more virtualizable resource groups (VRGs) for managing network traffic flow across a plurality of network switches on the network. Further, the RCAE is operable to add network entities to each VRG, add flows to each VRG, and add other VRGs to each VRG. A virtualizable resource control list (VRCL), associated with each VRG, identifies which network entities in the VRG can communicate with each other, which network entities in the VRG can communicate with network entities in other VRGs, and a guaranteed bandwidth for the VRG associated with the VRCL. Furthermore, the RCAE is operable to exchange messages with other RCAEs in other network switches to implement traffic policies defined by each VRCL.