Abstract: An addressing model is provided where all resources, including memory and devices, are addressed with internet protocol (IP) addresses. A task, such as an application, may be assigned a range of IP addresses rather than an effective address range. Thus, a processing element, such as an I/O adapter or even a printer, for example, may also be addressed using IP addresses without the need for library calls, device drivers, pinning memory, and so forth. This addressing model also provides full virtualization of resources across an IP interconnect, allowing a process to access an I/O device across a network.

Abstract: A heterogeneous processing element model is provided where I/O devices look and act like processors. In order to be treated like a processor, an I/O processing element, or other special purpose processing element, must follow some rules and have some characteristics of a processor, such as address translation, security, interrupt handling, and exception processing, for example. The heterogeneous processing element model abstracts an I/O device such that communication intended for the I/O device may be packetized and sent over a network. Thus, a virtualization platform may packetize communication intended for a remotely located I/O device and transmit the packetized communication over a distance, rather than having to make a call to a library, call a device driver, pin memory, and so forth.

Abstract: A mechanism for offloading the management of send queues in a split socket stack environment, including efficient split socket queue flow control and TCP/IP retransmission support. An Upper Layer Protocol (ULP) creates send work queue entries (SWQEs) for writing to the send work queue (SWQ). The Internet Protocol Suite Offload Engine (IPSOE) is notified of a new entry to the SWQ and it subsequently reads this entry that contains pointers to the data that is to be transmitted. After the data is transmitted and acknowledgments are received, the IPSOE creates a completion queue entry (CQE) that is written into the completion queue (CQ). The flow control between the ULP and the IPSOE is credit based. The passing of CQ credits is the only explicit mechanism required to manage flow control of both the SWQ and the CQ between the ULP and the IPSOE.

Abstract: A system and computer readable medium for oversubscribing bandwidth in a communication network, is disclosed. The system and computer readable medium includes policing a first data flow and outputting a first output data flow from the first meter, in relation to a first Committed Information Rate (CIR) and a first Peak Information Rate (PIR); policing a second data flow and outputting a second output data flow from the second meter in relation to a second CIR and a second PIR; and policing an aggregated output data flow of the first output data flow and the second output data through a third meter of the oversubscription module, where the aggregated output data flow is policed in relation to a third CIR and a third PIR.

Abstract: A technique for operating a high performance computing cluster includes monitoring workloads of multiple processors. The high performance computing cluster includes multiple nodes that each include two or more of the multiple processors. Workload information for the multiple processors is periodically updated in respective local job tables maintained in each of the multiple nodes. Based on the workload information in the respective local job tables, one or more threads are periodically moved to a different one of the multiple processors.

Abstract: A technique for operating a high performance computing (HPC) cluster includes monitoring workloads of multiple processors included in the HPC cluster. The HPC cluster includes multiple nodes that each include two or more of the multiple processors. One or more threads assigned to one or more of the multiple processors are moved to a different one of the multiple processors based on the workloads of the multiple processors.

Abstract: A technique for operating a high performance computing (HPC) cluster includes monitoring communication between threads assigned to multiple processors included in the HPC cluster. The HPC cluster includes multiple nodes that each include two or more of the multiple processors. One or more of the threads are moved to a different one of the multiple processors based on the communication between the threads.

Abstract: A technique for operating a high performance computing cluster (HPC) having multiple nodes (each of which include multiple processors) includes periodically broadcasting information, related to processor utilization and network utilization at each of the multiple nodes, from each of the multiple nodes to remaining ones of the multiple nodes. Respective local job tables maintained in each of the multiple nodes are updated based on the broadcast information. One or more threads are then moved from one or more of the multiple processors to a different one of the multiple processors (based on the broadcast information in the respective local job tables).

Abstract: An integrated processor design includes physical interface macros supporting heterogeneous electrical properties. The processor design comprises a plurality of processing cores and a plurality of physical interfaces to connect to a memory interface, a peripheral component interconnect express (PCI Express or PCIe) interface for input/output, an Ethernet interface for network communication, and/or a serial attached SCSI (SAS) interface for storage. Each physical interface may be programmatically connected to a selected interface controller, such as a memory controller, a PCI Express controller, or an Ethernet controller, for example. A plurality of such controllers may be connected to a switch within the processor design, with the switch also being connected to each physical interface macro. Thus, the physical interface macros may be programmatically connected to a subset of the plurality of controllers.

Abstract: Mechanisms for processing of communications between data processing devices are provided. With the mechanisms of the illustrative embodiments, a set of techniques that enables sustaining media speed by distributing transmit and receive-side processing over multiple processing cores is provided. In addition, these techniques also enable designing multi-threaded network interface controller (NIC) hardware that efficiently hides the latency of direct memory access (DMA) operations associated with data packet transfers over an input/output (I/O) bus. Multiple processing cores may operate concurrently using separate instances of a communication protocol stack and device drivers to process data packets for transmission with separate hardware implemented send queue managers in a network adapter processing these data packets for transmission.

Abstract: A method and apparatus are used for finding the longest prefix match in a variable length prefix search when searching a direct table within a routing table structure of a network processor. The search through the routing table structure is expedited by hashing a first segment of an internet protocol address with a virtual private network number followed by concatenating the unhashed bits of the IP address to the result of the hash operation to form an input key. Patterns are compared a bit at a time until an exact match or the best match is found. The search is conducted in a search tree that provides that the matching results will be the best possible match.

Abstract: Packet switching node in a communication system includes apparatus for receiving incoming information packets or frames which contain header portions with formatting control blocks. Information in the frame's header contains frame alteration commands for modifying the information in the frame. The modifications include adding new information, deleting information, and overlaying information. Decoders and control devices in an alteration engine interpret the commands and apply the modifications to the frame data. Common and standard data patterns are stored for insertion or overlaying to conserve data packet space.

Abstract: An addressing model is provided where devices, including I/O devices, are addressed with internet protocol (IP) addresses, which are considered part of the virtual address space. A task, such as an application, may be assigned an effective address range, which corresponds to addresses in the virtual address space. The virtual address space is expanded to include Internet protocol addresses. Thus, the page frame tables are also modified to include entries for IP addresses and additional properties for devices and I/O. Thus, a processing element, such as an I/O adapter or even a printer, for example, may also be addressed using IP addresses without the need for library calls, device drivers, pinning memory, and so forth. This addressing model also provides full virtualization of resources across an IP interconnect, allowing a process to access an I/O device across a network.

Abstract: A heterogeneous processing element model is provided where I/O devices look and act like processors. In order to be treated like a processor, an I/O processing element, or other special purpose processing element, must follow some rules and have some characteristics of a processor, such as address translation, security, interrupt handling, and exception processing, for example. The heterogeneous processing element model abstracts an I/O device such that communication intended for the I/O device may be packetized and sent over a network. Thus, a virtualization platform may packetize communication intended for a remotely located I/O device and transmit the packetized communication over a distance, rather than having to make a call to a library, call a device driver, pin memory, and so forth.

Abstract: An addressing model is provided where all resources, including memory and devices, are addressed with internet protocol (IP) addresses. A task, such as an application, may be assigned a range of IP addresses rather than an effective address range. Thus, a processing element, such as an I/O adapter or even a printer, for example, may also be addressed using IP addresses without the need for library calls, device drivers, pinning memory, and so forth. This addressing model also provides full virtualization of resources across an IP interconnect, allowing a process to access an I/O device across a network.

Abstract: A network processor dataflow chip and method for flexible dataflow are provided. The dataflow chip comprises a plurality of on-chip data transmission and scheduling circuit structures. The data transmission and scheduling circuit structures are selected responsive to indicators. Data transmission circuit structures may comprise selectable frame processing and data transmission functions. Selectable frame processing may comprise cut and paste, full dispatch and store and dispatch frame processing. Scheduling functions include full internal scheduling, calendar scheduling in communication with an external scheduler, and external calendar scheduling. In another aspect of the present invention, data transmission functions may comprise low latency and normal latency external processor interfaces for selectively providing privileged access to dataflow chip resources.

Abstract: Techniques and articles of manufacture are provided comprising computer readable programs that, when executed on the computer, cause the computer to delete a leaf from a Patricia tree having a direct table and a plurality of PSCB's which decode portions of the pattern of a leaf in the tree without shutting down the functioning of the tree. A leaf having a pattern is identified as a leaf to be deleted. Using the pattern, the tree is walked to identify the location of the leaf to be deleted. The leaf to be deleted is identified and deleted, and any relevant PSCB modified, if necessary. The technique also is applicable to deleting a prefix of a prefix.

Abstract: Apparatus and method for storing network frame data which is to be modified. A plurality of buffers stores the network data which is arranged in a data structure identified by a frame control block and buffer control block. A plurality of buffer control blocks associated with each buffer storing the frame data establishes a sequence of the buffers. Each buffer control block has data for identifying a subsequent buffer within the sequence. The first buffer is identified by a field of a frame control block as well as the beginning and ending address of the frame data. The frame data can be modified without rewriting the data to memory by altering the buffer control block and/or frame control block contents without having to copy or rewrite the data in order to modify it.

Abstract: A system and computer readable medium for oversubscribing bandwidth in a communication network, is disclosed. The system and computer readable medium includes policing a first data flow and outputting a first output data flow from the first meter, in relation to a first Committed Information Rate (CIR) and a first Peak Information Rate (PIR); policing a second data flow and outputting a second output data flow from the second meter in relation to a second CIR and a second PIR; and policing an aggregated output data flow of the first output data flow and the second output data through a third meter of the oversubscription module, where the aggregated output data flow is policed in relation to a third CIR and a third PIR.

Abstract: A mechanism for offloading the management of send queues in a split socket stack environment, including efficient split socket queue flow control and TCP/IP retransmission support. As consumers initiate send operations, send work queue entries (SWQEs) are created by an Upper Layer Protocol (ULP) and written to the send work queue (SWQ). The Internet Protocol Suite Offload Engine (IPSOE) is notified of a new entry to the SWQ and it subsequently reads this entry that contains pointers to the data that is to be transmitted. After the data is transmitted and acknowledgments are received, the IPSOE creates a completion queue entry (CQE) that is written into the completion queue (CQ). After the CQE is written, the ULP subsequently processes the entry and removes it from the CQE, freeing up a space in both the SWQ and CQ. The number of entries available in the SWQ are monitored by the ULP so that it does not overwrite any valid entries.