Abstract: The system (200) provides a photonic waveguide (210) formed on a substrate (220) and a plurality of steering mirrors (230) within the photonic waveguide. The steering mirrors can be configured to direct a light beam (240) between two or more computing components (260). A plurality of steering mirror supports (250) are located within the waveguide having preset locations. The steering mirror supports are configured to enable the steering mirrors to be selectively repositioned at the preset steering mirror supports within the photonic waveguide to create varying configurations. The steering mirrors in the varying configurations direct one or more optical beams to form multiple connectivity channels between computing components within the photonic waveguide.

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 method for connecting adjacent computing board devices. A source computing board may be provided. An optical engine attaches to the source computing board. A plurality of source optical connectors couples to the optical engine. A first optical connector may be positioned at a location on the source computing board for a first preset type of computing component on an adjacent computing board. A second optical connector may be positioned at a fixed coordinate related to the first optical connector on the source computing board.

Abstract: The present disclosure provides techniques for mapping large shared address spaces in a computing system. A method includes creating a physical address map for each node in a computing system. Each physical address map maps the memory of a node. Each physical address map is copied to a single address map to form a global address map that maps all memory of the computing system. The global address map is shared with all nodes in the computing system.

Abstract: A method for connecting adjacent computing board devices. A source computing board may be provided. An optical engine attaches to the source computing board. A plurality of source optical connectors couples to the optical engine. A first optical connector may be positioned at a location on the source computing board for a first preset type of computing component on an adjacent computing board. A second optical connector may be positioned at a fixed coordinate related to the first optical connector on the source computing board.

Abstract: A method of shielding a memory device (110) from high write rates comprising receiving instructions to write data at a memory container (105), the memory controller (105) composing a cache (120) comprising a number of cache lines defining stored data, with the memory controller (105), updating a cache line in response to a write hit in the cache (120), and with the memory controller (105), executing the instruction to write data in response to a cache miss to a cache line within the cache (120) in which the memory controller (105) prioritizes for writing to the cache (120) over writing to the memory device (110).

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 computer apparatus and related method to access storage is provided. In one aspect, a controller maps an address range of a data block of storage into an accessible memory address range of at least one of a plurality of processors, in a further aspect, the controller ensures that copies of the data block cached in a plurality of memories by a plurality of processors are consistent.

Abstract: In a computing system having a plurality of transaction source nodes issuing transactions into a switching fabric, an underserviced node notifies source nodes in the system that it needs additional system bandwidth to timely complete an ongoing transaction. The notified nodes continue to process already started transactions to completion, but stop the introduction of new traffic into the fabric until such time as the underserviced node indicates that it has progressed to a preselected point.

Abstract: Embodiments of the present invention are directed to high-bandwidth, low-latency optical fabrics for broadcasting between nodes. In one embodiment, an optical fabric includes an optical communication path optically coupled to a broadcasting node and optically coupled to one or more broadcast receiving nodes. The optical fabric also includes a first optical element optically coupled to the optical communication path and configured to broadcast an optical signal generated by the broadcasting nodes onto the optical communication path, and one or more optical elements optically coupled to the optical communication path and configured to divert a portion the broadcast optical signal onto each of the one or more receiving nodes.

Abstract: Various embodiments of the present invention are directed to arrangements of multiple optical buses to create scalable optical interconnect fabrics for computer systems. In one aspect, a multi-bus fabric (102) for transmitting optical signals between a plurality of nodes (108-111) comprises a plurality of optical buses (104-107). Each optical bus is optically coupled to each node of the plurality of nodes, and each optical bus is configured to so that one node broadcasts optical signals generated by the node to the other nodes of the plurality of nodes.

Abstract: A system (10) and methods (800, 900) for routing optical signals are disclosed. The system includes a large core hollow waveguide (30) having a reflective coating (40) covering an interior surface (32) of the hollow waveguide configured to guide a light beam (104). At least one area based beam splitter (50) is integrally formed with the hollow waveguide and has an angled reflective surface (52) with a selectable height (H) relative to the interior surface. The angled reflective surface is oriented to redirect a predetermined amount of the light beam (114) based on the height of the angled reflective surface.

Abstract: Feed-forward arbitration is disclosed. An example method of feed-forward arbitration includes determining an aggregated measure of urgency of packets waiting in a queue.

Abstract: An embodiment of a multiprocessor computer system comprises main memory, a remote processor capable of accessing the main memory, a remote cache device operative to store accesses by said remote processor to said main memory, and a filter tag cache device associated with the main memory. The filter cache device is operative to store information relating to remote ownership of data in the main memory including ownership by the remote processor. The filter cache device is operative to selectively invalidate filter tag cache entries when space is required in the filter tag cache device for new cache entries. The remote cache device is responsive to events indicating that a cache entry has low value to the remote processor to send a hint to the filter tag cache device. The filter tag cache device is responsive to a hint in selecting a filter tag cache entry to invalidate.

Abstract: Various embodiments of the present invention are directed to arrangements of multiple optical buses to create scalable optical interconnect fabrics for computer systems. In one aspect, a multi-bus fabric (102) for transmitting optical signals between a plurality of nodes (108-111) comprises a plurality of optical buses (104-107). Each optical bus is optically coupled to each node of the plurality of nodes, and each optical bus is configured to so that one node broadcasts optical signals generated by the node to the other nodes of the plurality of nodes.

Abstract: A method for connecting adjacent computing board devices. A source computing board may be provided. An optical engine attaches to the source computing board. A plurality of source optical connectors couples to the optical engine. A first optical connector may be positioned at a location on the source computing board for a first preset type of computing component on an adjacent computing board. A second optical connector may be positioned at a fixed coordinate related to the first optical connector on the source computing board.

Abstract: A system (10) and methods (800, 900) for routing optical signals are disclosed. The system includes a large core hollow waveguide (30) having a reflective coating (40) covering an interior surface (32) of the hollow waveguide configured to guide a light beam (104). At least one area based beam splitter (50) is integrally formed with the hollow waveguide and has an angled reflective surface (52) with a selectable height (H) relative to the interior surface. The angled reflective surface is oriented to redirect a predetermined amount of the light beam (114) based on the height of the angled reflective surface.

Abstract: In a computing system having a plurality of transaction source nodes issuing transactions into a switching fabric, an underserviced node notifies source nodes in the system that it needs additional system bandwidth to timely complete an ongoing transaction. The notified nodes continue to process already started transactions to completion, but stop the introduction of new traffic into the fabric until such time as the underserviced node indicates that it has progressed to a preselected point.

Abstract: Embodiments of the present invention are directed to high-bandwidth, low-latency optical fabrics for broadcasting between nodes. In one embodiment, an optical fabric includes an optical communication path optically coupled to a broadcasting node and optically coupled to one or more broadcast receiving nodes. The optical fabric also includes a first optical element optically coupled to the optical communication path and configured to broadcast an optical signal generated by the broadcasting nodes onto the optical communication path, and one or more optical elements optically coupled to the optical communication path and configured to divert a portion the broadcast optical signal onto each of the one or more receiving nodes.

Abstract: A computer system, related components such as a processor agent, and related method are disclosed. In at least one embodiment, the computer system includes a first core, at least one memory device including a first memory segment, and a first memory controller coupled to the first memory segment. Further, the computer system includes a fabric and at least one processor agent coupled at least indirectly to the first core and the first memory segment, and also coupled to the fabric. A first memory request of the first core in relation to a first memory location within the first memory segment proceeds to the first memory controller by way of the at least one processor agent and the fabric.