Abstract: Techniques are disclosed for reducing impact of a switch failure and/or a repair action in a switch fabric. In one embodiment, a server system is provided that includes a first interposer card that operatively connects one or more server cards to a midplane. The first interposer card may include a switch module that switches network traffic for the one or more server cards. The first interposer card may be hot-swappable from the midplane, and the one or more server cards may be hot-swappable from the first interposer card. The server system may further include an interconnect between the first interposer card and a second interposer card.

Abstract: Method and apparatus for optimally placing memory devices within a computer system. A memory controller may include circuitry configured to retrieve or one or more performance metrics a plurality of memory devices connected thereto. Based on the performance metrics and one or more predefined rules for placing memory devices, the circuitry may determine an optimal placement of the memory devices in the system.

Abstract: Techniques are disclosed for managing a switch fabric. In one embodiment, a server system is provided that includes a midplane, one or more server cards, switch modules and a management controller. The midplane may include a fabric interconnect for a switch fabric. The one or more server cards and the switch modules may be operatively connected to the midplane. The switch modules may be configured to switch network traffic for the one or more server cards. The management controller may be configured to manage the switch modules via the fabric interconnect.

Abstract: A method and system are provided for implementing enhanced memory performance management with configurable bandwidth versus power usage in a chip stack of memory chips. A chip stack of memory chips is connected in a predefined density to allow a predefined high bandwidth connection between each chip in the stack, such as with through silicon via (TSV) interconnections. Large-bandwidth data transfers are enabled from the memory chip stack by trading off increased power usage for memory performance on a temporary basis.

Abstract: Systems and methods to manage memory are provided. A particular method may include initiating a memory compression operation. The method may further include initiating a first interrupt configured to affect a first process executing on a processor in response to a first detected memory level. A second initiated interrupt may be configured to affect the first process executing on the processor in response to a second detected memory level, and a third interrupt may be initiated to affect the first process executing on the processor in response to a third detected memory level. At least of the first, the second, and the third detected memory levels are affected by the memory compression operation.

Abstract: Techniques are disclosed for managing a switch fabric. In one embodiment, a server system is provided that includes a midplane, one or more server cards, switch modules and a management controller. The midplane may include a fabric interconnect for a switch fabric. The one or more server cards and the switch modules may be operatively connected to the midplane. The switch modules may be configured to switch network traffic for the one or more server cards. The management controller may be configured to manage the switch modules via the fabric interconnect.

Abstract: Techniques are disclosed for reducing impact of a switch failure and/or a repair action in a switch fabric. In one embodiment, a server system is provided that includes a first interposer card that operatively connects one or more server cards to a midplane. The first interposer card may include a switch module that switches network traffic for the one or more server cards. The first interposer card may be hot-swappable from the midplane, and the one or more server cards may be hot-swappable from the first interposer card. The server system may further include an interconnect between the first interposer card and a second interposer card.

Abstract: Techniques are disclosed for reducing impact of a repair action in a switch fabric. In one embodiment, a server system is provided that includes a first interposer card that operatively connects one or more server cards to a midplane. The first interposer card may include a switch module that switches network traffic for the one or more server cards. The first interposer card may be hot-swappable from the midplane, and the one or more server cards may be hot-swappable from the first interposer card.

Abstract: Techniques are disclosed for reducing impact of a switch failure in a switch fabric. In one embodiment, a server system is provided that includes a midplane, one or more server cards and one or more switch cards. The midplane may include a fabric interconnect for a switch fabric. The one or more server cards may be coupled with the midplane, where each server card is hot-swappable from the midplane. The one or more switch cards may also be coupled with the midplane, where each switch card is also hot-swappable from the midplane. Each switch card includes one or more switch modules, and each switch module is configured to switch network traffic for at least one server card.

Abstract: A design structure for controlling computer-readable memory includes a plurality of memory locations, a usage frequency of a data unit stored in a first memory location is determined. The data unit is moved to a second memory location, different from the first memory location that is selected based on a correspondence between a known latency of the second memory location and the usage frequency of the data unit, in which the second memory location is the primary data storage location for the data unit.

Abstract: A computer program product and computer implemented method are provided for migrating a virtual machine between servers. The virtual machine is initially operated on a first server, wherein the first server accesses the virtual machine image over a network at a memory location within fabric attached memory. The virtual machine is migrated from the first server to a second server by flushing data to the virtual machine image from cache memory associated with the virtual machine on the first server and providing the state and memory location of the virtual machine to the second server. The virtual machine may then operate on the second server, wherein the second server accesses the virtual machine image over the network at the same memory location within the fabric attached memory without copying the virtual machine image.

Abstract: Systems and methods to manage memory are provided. A particular method may include selecting one of a plurality of compression modes to perform memory compression operations at a server computer. The plurality of compression modes may include a first memory compression mode configured to perform a first memory compression operation using a compression engine, and a second compression mode configured to perform a second memory compression operation using the compression engine. At least one of the first compression operation and the second compression operation may be performed according to the selected compression mode.

Abstract: Systems and methods to manage memory are provided. A particular method may include initiating a memory compression operation. The method may further include initiating a first interrupt configured to affect a first process executing on a processor in response to a first detected memory level. A second initiated interrupt may be configured to affect the first process executing on the processor in response to a second detected memory level, and a third interrupt may be initiated to affect the first process executing on the processor in response to a third detected memory level. At least of the first, the second, and the third detected memory levels are affected by the memory compression operation.

Abstract: A system includes a processor providing hardware virtualization, and a memory to store a first virtual machine data structure corresponding to a first nested virtualization level and a second virtual machine data structure corresponding to a second nested virtualization level. The virtual machine data structures assist management of the hardware virtualization provided by the processor. The system includes a first nested virtual machine located within the first nested virtualization level and a second nested virtual machine located within the second nested virtualization level. The system includes hypervisors to manage the nested virtual machines using the virtual machine data structures. A root hypervisor is to manage the first nested virtual machine using the first virtual machine data structure. A first nested virtualization level hypervisor is to run within the first nested virtual machine and is to manage the second nested virtual machine using the second virtual machine data structure.

Abstract: Systems and methods to manage memory are provided. A particular method may include storing data in a primary memory that is in communication with a processor and storing in a mirrored data in a mirrored memory. The mirrored data may be compressed, and the mirrored memory may be in communication with the processor. A failure condition associated with the data of the primary memory may be detected. In response to the detected failure condition, the mirrored data in the mirrored memory may be accessed.

Abstract: A design structure is provided for a dual-mode memory chip supporting a first operation mode in which received data access commands contain chip select data to identify the chip addressed by the command, and control logic in the memory chip determines whether the command is addressed to the chip, and a second operation mode in which the received data access command addresses a set of multiple chips. Preferably, the first mode supports a daisy-chained configuration of memory chips. Preferably the second mode supports a hierarchical interleaved memory subsystem, in which each addressable set of chips is configured as a tree, command and write data being propagated down the tree, the number of chips increasing at each succeeding level of the tree.

Abstract: A method, an apparatus and a program product may enable scalable bandwidth and memory for a system having processor with directly attached memory. Multiple memory expander microchips may include non-volatile memory to provide additional memory bandwidth and capacity while in communication with the processor. The uncompressed data region may be implemented with standard high speed dynamic random access memory. The less frequently accessed compressed data region may be implemented with non-volatile memory to leverage its benefits of higher density, more capacity, and lower power compared to DRAM. Memory and bandwidth allocation between may be dynamically adjusted.

Abstract: A design structure is provided for a memory module containing an interface for receiving memory access commands from an external source, in which a first portion of the interface receives memory access data at a first bus frequency and a second portion of the interface receives memory access data at a second different bus frequency. Preferably, the memory module contains a second interface for re-transmitting memory access data, also operating at dual frequency. The memory module is preferably used in a high-capacity memory subsystem organized in a tree configuration in which data accesses are interleaved. Preferably, the memory module has multiple-mode operation, one of which supports dual-speed buses for receiving and re-transmitting different parts of data access commands, and another of which supports conventional daisy-chaining.

Abstract: A design structure is provided for a memory module containing a first interface for receiving data access commands and a second interface for re-transmitting data access commands to other memory modules, the second interface propagating multiple copies of received data access commands to multiple other memory modules. The memory module is preferably used in a high-capacity memory subsystem organized in a tree configuration in which data accesses are interleaved. Preferably, the memory module has multiple-mode operation, one of which supports multiple replication of commands and another of which supports conventional daisy-chaining.

Abstract: Method and apparatus for optimally placing memory devices within a computer system. A memory controller may include circuitry configured to retrieve or one or more performance metrics a plurality of memory devices connected thereto. Based on the performance metrics and one or more predefined rules for placing memory devices, the circuitry may determine an optimal placement of the memory devices in the system.