In response to determining circuitry determining that a portion of data to be sent to a recipient over an interconnect has a predetermined value, data sending circuitry performs data elision to: omit sending at least one data FLIT corresponding to the portion of data having the predetermined value; and send a data-elision-specifying FLIT specifying data-elision information indicating to the recipient that sending of the at least one data FLIT has been omitted and that the recipient can proceed assuming the portion of data has the predetermined value. The data-elision-specifying FLIT is a FLIT other than a write request FLIT for initiating a memory write transaction sequence. This helps to conserve data FLIT bandwidth for other data not having the predetermined value.

A processor-to-processor agent to provide connectivity over a processor-to-processor interconnect between services/network functions on different processors on a same compute node in a server is provided. The processor-to-processor agent can intercept socket interface calls using a network traffic filter in the network stack and redirect the packets based on traffic matching rules.

Embodiments herein describe end-to-end bindings to create zones that extend between different components in a SoC, such as an I/O gateway, a processor subsystem, a NoC, storage and data accelerators, programmable logic, etc. Each zone can be assigned to a different domain that is controlled by a tenant such as an external host, or software executing on that host. Embodiments herein create end-to-end bindings between acceleration engines, I/O gateways, and embedded cores in SoCs. Instead of these components being treated as disparate monolithic components, the bindings divide up the hardware and memory resources across components that make up the SoC, into different zones. Those zones in turn can have unique bindings to multiple tenants. The bindings can be configured in bridges between components to divide resources into the zones to enable tenants of those zones to have dedicated available resources that are secure from the other tenants.

An interface for coupling an agent to a fabric supports a set of coherent interconnect protocols and includes a global channel to communicate control signals to support the interface, a request channel to communicate messages associated with requests to other agents on the fabric, a response channel to communicate responses to other agents on the fabric, and a data channel to couple to communicate messages associated with data transfers to other agents on the fabric, where the data transfers include payload data.

A multi-chip module includes a first Integrated Circuit (IC) die a second IC die. The first IC die includes an array of first bond pads, a plurality of first code group circuits, and first interleaved interconnections between the plurality of first code group circuits and the array of first bond pads, the first interleaved interconnections including a first interleaving pattern causing data from different code group circuits to be coupled to adjacent first bond pads. The second IC die includes a second array of bond pads that electrically couple to the array of first bond pads, a plurality of second code group circuits, and second interleaved interconnections between the plurality of second code group circuits and the array of second bond pads, the second interleaved interconnections including a second interleaving pattern causing data from different code groups to be coupled to adjacent second bond pads.

Packet routing within a multi-chip processing system is shown. A first chip has a first interconnect bus, and a first microprocessor coupled to the first interconnect bus. The first interconnect bus has a first routing register. When the first microprocessor operates the first chip as a source node to output a packet to be transferred to a destination node, routing information indicating a routing path from the source node to the destination node is written into the first routing register and then loaded from the first routing register to a header of the packet. While being transferred within the multi-chip processing system from the source node to the destination node, the packet is guided along the routing path indicated in the routing information carried in the header of the packet.

A Peripheral Component Interconnect Express (PCIe)-based data transmission method and apparatus includes a first node that encapsulates data into a transaction layer packet (TLP) and then sends the TLP to a second node, where the TLP includes a packet header part, a first field and a second field of the packet header part that are used to indicate first encapsulation information, and the first encapsulation information includes a data type of the data and at least one encapsulation parameter corresponding to the data type.t The first field and the second field are used to indicate the information required for transmitting the data, so that the endpoints can communicate with each other even if the root is not used.

Methods and systems are provided for performing lossy dropping and ECN marking in a flow-based network. The system can maintain state information of individual packet flows, which can be set up or released dynamically based on injected data. Each flow can be provided with a flow-specific input queue upon arriving at a switch. Packets of a respective flow are acknowledged after reaching the egress point of the network, and the acknowledgement packets are sent back to the ingress point of the flow along the same data path. As a result, each switch can obtain state information of each flow and perform per-flow packet dropping and ECN marking.

This application discloses a peripheral component interconnect express (PCIe)-based data transmission method and apparatus. The method includes: A first node encapsulates data into a transaction layer packet (TLP) and then sends the TLP to a second node. The TLP includes a packet header and an extension header. The packet header includes a first field and a second field. The first field, the second field, and the extension header are used to indicate first encapsulation information. The first encapsulation information includes a data type of the data and at least one encapsulation parameter corresponding to the data type. In some embodiments, the first field, the second field, and the extension header are used to indicate the information required for transmitting the data.

There is disclosed an example of an artificial intelligence (AI) system, including: a first hardware platform; a fabric interface configured to communicatively couple the first hardware platform to a second hardware platform; a processor hosted on the first hardware platform and programmed to operate on an AI problem; and a first training accelerator, including: an accelerator hardware; a platform inter-chip link (ICL) configured to communicatively couple the first training accelerator to a second training accelerator on the first hardware platform without aid of the processor; a fabric ICL to communicatively couple the first training accelerator to a third training accelerator on a second hardware platform without aid of the processor; and a system decoder configured to operate the fabric ICL and platform ICL to share data of the accelerator hardware between the first training accelerator and second and third training accelerators without aid of the processor.