A network system for a data center is described in which a switch fabric provides full mesh interconnectivity such that any servers may communicate packet data to any other of the servers using any of a number of parallel data paths. Moreover, according to the techniques described herein, edge-positioned access nodes, optical permutation devices and core switches of the switch fabric may be configured and arranged in a way such that the parallel data paths provide single L2/L3hop, full mesh interconnections between any pairwise combination of the access nodes, even in massive data centers having tens of thousands of servers. The plurality of optical permutation devices permute communications across the optical ports based on wavelength so as to provide full-mesh optical connectivity between edge-facing ports and core-facing ports without optical interference.

This disclosure describes techniques for performing communications between devices using various aspects of Ethernet standards. As further described herein, a protocol is disclosed that may be used for communications between devices, where the communications take place over a physical connection complying with Ethernet standards. Such a protocol may enable reliable and in-order delivery of frames between devices, while following Ethernet physical layer rules, Ethernet symbol encoding, Ethernet lane alignment, and/or Ethernet frame formats.

A network system includes P upper switches, Q lower switches, and a first mapping device. Each upper switch of the P upper switches includes a plurality of upper ports. A group of upper switches selected from the P upper switches includes P1 upper switches. Each lower switch of the Q lower switches includes a plurality of upper ports. The first mapping device includes P1 upper adapter terminals coupled to a part of upper ports of the P1 upper switches, and P1 lower adapter terminals coupled to lower ports of a part of Q lower switches. The first mapping device is used for allocating a plurality of transmitting channels and receiving channels received by each upper adapter terminal to the P1 lower adapter terminals.

A network system includes P upper switches, Q lower switches, and a first mapping device. Each upper switch of the P upper switches includes a plurality of upper ports. A group of upper switches selected from the P upper switches includes P1 upper switches. Each lower switch of the Q lower switches includes a plurality of upper ports. The first mapping device includes P1 upper adapter terminals coupled to a part of upper ports of the P1 upper switches, and P1 lower adapter terminals coupled to lower ports of a part of Q lower switches. The first mapping device is used for allocating a plurality of transmitting channels and receiving channels received by each upper adapter terminal to the P1 lower adapter terminals.

An information transmission method and a related device applied to a cloud service system are provided. The cloud service system includes an arbiter, a first switch, and a first server group. The arbiter is connected to the first server group by using the first switch. The method includes: a first server obtains a first identifier of the first server in the first server group. The first server determines, in a first slot, a first time domain location corresponding to the first identifier. The first time domain location is a time domain location at which the first switch receives a first request to be processed by the arbiter, and different identifiers correspond to different time domain locations in the first slot. The first server sends the first request to the first switch based on the first time domain location.

An information transmission method and a related device applied to a cloud service system are provided. The cloud service system includes an arbiter, a first switch, and a first server group. The arbiter is connected to the first server group by using the first switch. The method includes: a first server obtains a first identifier of the first server in the first server group. The first server determines, in a first slot, a first time domain location corresponding to the first identifier. The first time domain location is a time domain location at which the first switch receives a first request to be processed by the arbiter, and different identifiers correspond to different time domain locations in the first slot. The first server sends the first request to the first switch based on the first time domain location.

In some embodiments, a data packet may be received at a leaf switch. A port-channel associated with a destination port for the data packet may be identified, and the data packet may be transmitted to the destination port via the identified port-channel.

A method of operation of a computing system includes: providing a first cluster having a first kernel unit for managing a first reconfigurable hardware device; analyzing an application descriptor associated with an application; generating a first bitstream based on the application descriptor for loading the first reconfigurable hardware device, the first bitstream for implementing at least a first portion of the application; and implementing a first fragment with the first bitstream in the first cluster.

A distributed switch architecture supports very high bandwidth applications. For instance, the distributed switch architecture may be implemented for cloud networks. The architecture scales by organizing traffic management components into tiled structures with distributed buffering. The tile structures are replicated and interconnected to perform transfers from ingress to egress using an interconnect bandwidth scheduling algorithm. Bandwidth scaling may be achieved by adding more tiles to achieve higher bandwidth. The interconnect in the architecture may be swapped out depending on implementation parameters, e.g., physical efficiency.

A distributed switch architecture supports very high bandwidth applications. For instance, the distributed switch architecture may be implemented for cloud networks. The architecture scales by organizing traffic management components into tiled structures with distributed buffering. The tile structures are replicated and interconnected to perform transfers from ingress to egress using an interconnect bandwidth scheduling algorithm. Bandwidth scaling may be achieved by adding more tiles to achieve higher bandwidth. The interconnect in the architecture may be swapped out depending on implementation parameters, e.g., physical efficiency.