Abstract: A system for hot swapping a network switch without disconnecting the network switch connectors is provided. The system disaggregates the switch faceplate network cable connectors from the internal components of the network switch so that the internal switch components may be removed from the switch without disconnecting the switch network cables.

Abstract: A co-packaged, multiplane network includes: an enclosure; a portion of a first network plane disposed within the enclosure and comprising a first plurality of interconnected switches; a portion of a second network plane disposed within the enclosure and comprising a second plurality of interconnected switches, the second network plane being independent of the first network plane and having the same topology as the first network plane; and a plurality of connectors, each connector being communicatively coupled to a respective port of each of the first plurality of interconnected switches and the second plurality of interconnected switches.

Abstract: A system for hot swapping a network switch without disconnecting the network switch connectors is provided. The system disaggregates the switch faceplate network cable connectors from the internal components of the network switch so that the internal switch components may be removed from the switch without disconnecting the switch network cables.

Abstract: A technique includes, in response to a cache miss occurring with a given processing node of a plurality of processing nodes, using a directory-based coherence system for the plurality of processing nodes to regulate snooping of an address that is associated with the cache miss. Using the directory-based coherence system to regulate whether the address is included in a snooping domain is based at least in part on a number of cache misses associated with the address.

Abstract: A system for hot swapping a network switch without disconnecting the network switch connectors is provided. The system disaggregates the switch faceplate network cable connectors from the internal components of the network switch so that the internal switch components may be removed from the switch without disconnecting the switch network cables.

Abstract: A computing system is disclosed herein. The computing system includes a computing node and a remote memory node coupled to the computing node via a system fabric. The computing node includes a plurality of processors and a master memory controller. The master memory controller is external to the plurality of processors. The master memory controller routes requests corresponding to requests from the plurality of processors across the system fabric to the remote memory node and returns a response.

Abstract: Techniques are provided for overcoming failures in a memory. One portion of the memory may operate in a single chip spare mode. Upon detection of an error in a single chip in the portion of the memory, a region of the portion of the memory may be converted to operate in a double chip spare mode. The memory may be accessed in both single and double chip spare modes.

Abstract: A co-packaged, multiplane network includes: an enclosure; a portion of a first network plane disposed within the enclosure and comprising a first plurality of interconnected switches; a portion of a second network plane disposed within the enclosure and comprising a second plurality of interconnected switches, the second network plane being independent of the first network plane and having the same topology as the first network plane; and a plurality of connectors, each connector being communicatively coupled to a respective port of each of the first plurality of interconnected switches and the second plurality of interconnected switches.

Abstract: According to an example, a lock may be requested by a first redundancy controller from a parity media controller to perform a first sequence that accesses multiple memory modules in a stripe. The lock may be acquired for the stripe so that the first sequence may be performed on the stripe. The lock may then be released from the stripe.

Abstract: A technique includes, in response to a cache miss occurring with a given processing node of a plurality of processing nodes, using a directory-based coherence system for the plurality of processing nodes to regulate snooping of an address that is associated with the cache miss. Using the directory-based coherence system to regulate whether the address is included in a snooping domain is based at least in part on a number of cache misses associated with the address.

Abstract: A technique includes, in response to a cache miss occurring with a given processing node of a plurality of processing nodes, using a directory-based coherence system for the plurality of processing nodes to regulate snooping of an address that is associated with the cache miss. Using the directory-based coherence system to regulate whether the address is included in a snooping domain is based at least in part on a number of cache misses associated with the address.

Abstract: According to an example, a lock may be requested by a first redundancy controller from a parity media controller to perform a first sequence that accesses multiple memory modules in a stripe. The lock may be acquired for the stripe so that the first sequence may be performed on the stripe. The lock may then be released from the stripe.

Abstract: A PCIe (Peripheral Component Interconnect Express) protocol converter for connection to a central processing unit (CPU) node having a root complex, a CPU memory fabric and CPU memory may include independent PCIe links, a fabric interface and a fabric switch connected to the fabric interface. Each of the links may include an endpoint for connection to the root complex. The fabric switch is connected to the fabric interface of each of the links and is connectable to a remote node. The fabric switch transmits writes of a single write request from the remote node across both links. Each fabric interface is to transmit an acknowledgment to the remote node in response to a write of the writes becoming observable at the CPU node hi Michael, hi Michael,.

Abstract: A computing system is disclosed herein. The computing system includes a computing node and a remote memory node coupled to the computing node via a system fabric. The computing node includes a plurality of processors and a master memory controller. The master memory controller is external to the plurality of processors. The master memory controller routes requests corresponding to requests from the plurality of processors across the system fabric to the remote memory node and returns a response.

Abstract: In some examples, each processor of a plurality of processors applies an interleave transform to perform interleaved access of a plurality of memory banks, where for any given memory address in use by the plurality of processors, applying any of the interleave transforms results in selection of a same memory bank of the plurality of memory banks and a same address within the same memory bank.

Abstract: According to an example, a lock may be requested by a first redundancy controller from a parity media controller to perform a first sequence that accesses multiple memory modules in a stripe. The lock may be acquired for the stripe so that the first sequence may be performed on the stripe. The lock may then be released from the stripe.

Abstract: Examples relate to virtual channel routing in networks considering VC actions to be performed by the packets while routed through the network. A packet is received at an input port of a network device of a network and an output port and a VC action is determined from a routing table associated to the input port based on a packet's destination network device. A VC mask is determined from a Virtual Channel Action Table (VCAT), associated to the routing table, based on a packet's ingress VC and the VC action. A particular VC among the set of VCs defined in the VC mask is selected and the packet is routed to the destination network device using the output port and the particular VC.

Abstract: Techniques are provided for overcoming failures in a memory. One portion of the memory may operate in a single chip spare mode. Upon detection of an error in a single chip in the portion of the memory, a region of the portion of the memory may be converted to operate in a double chip spare mode. The memory may be accessed in both single and double chip spare modes.

Abstract: A computing system is disclosed herein. The computing system includes a computing node and a remote memory node coupled to the computing node via a system fabric. The computing node includes a plurality of processors and a master memory controller. The master memory controller is external to the plurality of processors. The master memory controller routes requests corresponding to requests from the plurality of processors across the system fabric to the remote memory node and returns a response.

Abstract: A computing system can include a processor and a persistent main memory including a fault tolerance capability. The computing system can also include a memory controller to store data in the persistent main memory and create redundant data. The memory controller can also store the redundant data remotely with respect to the persistent main memory. The memory controller can further access the redundant data during failure of the persistent main memory.