Abstract: Embodiments relate to first-in-first-out (FIFO) queue based command spreading. An aspect includes receiving a plurality of commands by a first level priority stage of a memory control unit (MCU), wherein each of the plurality of commands is associated with one of a plurality of ports located on a buffer chip. Another aspect includes storing each of the plurality of commands in a FIFO queue of a plurality of FIFO queues in the MCU, wherein each of the plurality of commands is assigned to a FIFO queue based on the command's associated port, and each of the plurality of FIFO queues is associated with a respective one of the plurality of ports located on the buffer chip. Another aspect includes selecting a FIFO queue of the plurality of FIFO queues and forwarding a command from the selected FIFO queue to the buffer chip by the second level priority stage.

Abstract: Embodiments relate to reestablishing synchronization across multiple channels in a memory system. One aspect is a system that includes a plurality of channels, each providing communication with a memory buffer chip and a plurality of memory devices. A memory control unit is coupled to the plurality of channels. The memory control unit is configured to perform a method that includes receiving an out-of-synchronization indication associated with at least one of the channels. The memory control unit performs a first stage of reestablishing synchronization that includes selectively stopping new traffic on the plurality of channels, waiting for a first time period to expire, resuming traffic on the plurality of channels based on the first time period expiring, and verifying that synchronization is reestablished for a second time period.

Abstract: Embodiments relate to a dual asynchronous and synchronous memory system. One aspect is a system that includes a memory controller and a memory buffer chip coupled to the memory controller via a synchronous channel. The memory buffer chip includes a memory buffer unit configured to synchronously communicate with the memory controller in a nest domain, and a memory buffer adaptor configured to communicate with at least one memory interface port in a memory domain. The at least one memory interface port is operable to access at least one memory device. A boundary layer is connected to the nest domain and the memory domain, where the boundary layer is configurable to operate in a synchronous transfer mode between the nest and memory domains and to operate in an asynchronous transfer mode between the nest and memory domains.

Abstract: Embodiments relate to address mapping including generic bits. An aspect includes receiving an address including generic bits from a memory control unit (MCU) by a buffer module in a main memory. Another aspect includes mapping the generic bits to an address format corresponding to a type of dynamic random access memory (DRAM) in a memory subsystem associated with the buffer module by the buffer module. Yet another aspect includes accessing a physical location in the DRAM in the memory subsystem by the buffer module based on the mapped generic bits.

Abstract: Embodiments relate to first-in-first-out (FIFO) queue based command spreading. An aspect includes receiving a plurality of commands by a first level priority stage of a memory control unit (MCU), wherein each of the plurality of commands is associated with one of a plurality of ports located on a buffer chip. Another aspect includes storing each of the plurality of commands in a FIFO queue of a plurality of FIFO queues in the MCU, wherein each of the plurality of commands is assigned to a FIFO queue based on the command's associated port, and each of the plurality of FIFO queues is associated with a respective one of the plurality of ports located on the buffer chip. Another aspect includes selecting a FIFO queue of the plurality of FIFO queues and forwarding a command from the selected FIFO queue to the buffer chip by the second level priority stage.

Abstract: Providing heterogeneous recovery in a redundant memory system that includes a memory controller, a plurality of memory channels in communication with the memory controller, an error detection code mechanism configured for detecting a failing memory channel, and an error recovery mechanism. The error recovery mechanism is configured for receiving notification of the failing memory channel, for performing a recovery operation on the failing memory channel while other memory channels are performing normal system operations, for bringing the recovered channel back into operational mode with the other memory channels for store operations, for continuing to mark the recovered channel to guard against stale data, for removing any stale data after the recovery operation is complete, and for removing the mark on the recovered channel to allow the normal system operations with all of the memory channels, the removing based on the removing any stale data being complete.

Abstract: A computer implemented method for providing homogeneous recovery in a redundant memory system. The method includes receiving a notification that a memory channel has failed, where the memory channel is one of a plurality of memory channels in a memory system. New operations are blocked from starting on the memory channels in response to the notification, and any pending operations on the memory channels are completed in response to the notification. A recovery operation is performed on the memory channels in response to the completing. The new operations are started on at least a first subset of the memory channels in response to the recovery operation completing. The memory system is configured to operate with the first subset of the memory channels.

Abstract: Maintaining cache coherence in a multi-node, symmetric multiprocessing computer, the computer composed of a plurality of compute nodes, including, broadcasting upon a cache miss by the first compute node to other compute nodes a request for the cache line; if at least two of the compute nodes has a correct copy of the cache line, selecting which compute node is to transmit the correct copy of the cache line to the first node, and transmitting from the selected compute node to the first node the correct copy of the cache line; and updating by each node the state of the cache line in each node, in dependence upon one or more of the states of the cache line in all the nodes.

Abstract: Embodiments relate to storing data to a system memory. An aspect includes accessing successive entries of a cache directory having a plurality of directory entries by a stepper engine, where access to the cache directory is given a lower priority than other cache operations. It is determined that a specific directory entry in the cache directory has a change line state that indicates it is modified. A store operation is performed to send a copy of the specific corresponding cache entry to the system memory as part of a cache management function. The specific directory entry is updated to indicate that the change line state is unmodified.

Abstract: Providing heterogeneous recovery in a redundant memory system that includes a memory controller, a plurality of memory channels in communication with the memory controller, an error detection code mechanism configured for detecting a failing memory channel, and an error recovery mechanism. The error recovery mechanism is configured for receiving notification of the failing memory channel, for performing a recovery operation on the failing memory channel while other memory channels are performing normal system operations, for bringing the recovered channel back into operational mode with the other memory channels for store operations, for continuing to mark the recovered channel to guard against stale data, for removing any stale data after the recovery operation is complete, and for removing the mark on the recovered channel to allow the normal system operations with all of the memory channels, the removing in response to the removing any stale data being complete.

Abstract: Embodiments relate to storing data to a system memory. An aspect includes accessing successive entries of a cache directory having a plurality of directory entries by a stepper engine, where access to the cache directory is given a lower priority than other cache operations. It is determined that a specific directory entry in the cache directory has a change line state that indicates it is modified. A store operation is performed to send a copy of the specific corresponding cache entry to the system memory as part of a cache management function. The specific directory entry is updated to indicate that the change line state is unmodified.

Abstract: A method and apparatus for eliminating, in a multi-nodes data handling system, contention for exclusivity of lines in cache memory through improved management of system buses, processor cross-invalidate stacks, and the system operations that can lead to these requested cache operations being rejected.

Abstract: Maintaining cache coherence in a multi-node, symmetric multiprocessing computer, the computer composed of a plurality of compute nodes, including, broadcasting upon a cache miss by a first compute node a request for a cache line; transmitting from each of the other compute nodes to all other nodes the state of the cache line on that node, including transmitting from any compute node having a correct copy to the first node the correct copy of the cache line; and updating by each node the state of the cache line in each node, in dependence upon one or more of the states of the cache line in all the nodes.

Abstract: The method includes initiating a processor request to a cache in a requesting node and broadcasting the processor request to remote nodes when the processor request encounters a local cache miss, performing a directory search of each remote cache to determine a state of a target line's address and an ownership state of a specified address, returning the state of the target line to the requesting node and forming a combined response, and broadcasting the combined response to each remote node. During a fetch operation, when the directory search indicates an IM or a Target Memory node on a remote node, data is sourced from the respective remote cache and forwarded to the requesting node while protecting the data, and during a store operation, the data is sourced from the requesting node and protected while being forwarded to the IM or the Target Memory node after coherency has been established.

Abstract: Improved access to retained data useful to a system is accomplished by managing data flow through cache associated with the processor(s) of a multi-node system. A data management facility operable with the processors and memory array directs the flow of data from the processors to the memory array by determining the path along which data evicted from a level of cache close to one of the processors is to return to a main memory and directing evicted data to be stored, if possible, in a horizontally associated cache.

Abstract: Providing heterogeneous recovery in a redundant memory system that includes a memory controller, a plurality of memory channels in communication with the memory controller, an error detection code mechanism configured for detecting a failing memory channel, and an error recovery mechanism. The error recovery mechanism is configured for receiving notification of the failing memory channel, for performing a recovery operation on the failing memory channel while other memory channels are performing normal system operations, for bringing the recovered channel back into operational mode with the other memory channels for store operations, for continuing to mark the recovered channel to guard against stale data, for removing any stale data after the recovery operation is complete, and for removing the mark on the recovered channel to allow the normal system operations with all of the memory channels, the removing in response to the removing any stale data being complete.

Abstract: Providing homogeneous recovery in a redundant memory system that includes a memory controller, a plurality of memory channels in communication with the memory controller, an error detection code mechanism configured for detecting a failing memory channel, and an error recovery mechanism. The error recovery mechanism is configured for receiving notification of the failing memory channel, for blocking off new operations from starting on the memory channels, for completing any pending operations on the memory channels, for performing a recovery operation on the memory channels and for starting the new operations on at least a first subset of the memory channels. The memory system is capable of operating with the first subset of the memory channels.

Abstract: Maintaining cache coherence in a multi-node, symmetric multiprocessing computer, the computer composed of a plurality of compute nodes, including, broadcasting upon a cache miss by the first compute node to other compute nodes a request for the cache line; if at least two of the compute nodes has a correct copy of the cache line, selecting which compute node is to transmit the correct copy of the cache line to the first node, and transmitting from the selected compute node to the first node the correct copy of the cache line; and updating by each node the state of the cache line in each node, in dependence upon one or more of the states of the cache line in all the nodes.

Abstract: Maintaining cache coherence in a multi-node, symmetric multiprocessing computer, the computer composed of a plurality of compute nodes, including, broadcasting upon a cache miss by a first compute node a request for a cache line; transmitting from each of the other compute nodes to all other nodes the state of the cache line on that node, including transmitting from any compute node having a correct copy to the first node the correct copy of the cache line; and updating by each node the state of the cache line in each node, in dependence upon one or more of the states of the cache line in all the nodes.

Abstract: A method and apparatus in which the observability of cross-invalidates requests within remote nodes is controlled at the time of a partial response generation, when a remote request initially checks/snoops the directory state of the remote node, but before such the time that the cross-invalidate request is actually sent to the processors on a given node. If all of the remote nodes in the system indicate that the cross-invalidates could be sent during an initial directory snoop, the requesting node is able to return full exclusivity to a given cache line to a requesting processor at the time when it receives all of the partial responses, instead of having to wait for the final responses from each of the remote nodes within the system.