Abstract: Apparatus selectively attachable to the outer case of a cellular telephone or other similarly sized portable device. The apparatus desirably includes a housing containing an elastomeric band that is principally disposed inside the housing when the apparatus is not in use. The apparatus may include a kickstand and a cardholder. The elastomeric band can be used as a selectively retractable carrying strap when the cell phone or other similar device is not in use. Alternatively, the flexible continuous loop can be used, for example, to support the cell phone around a user's neck for hands-free use when the user is engaged in other activities.

Abstract: A memory controller that performs address replacement using combinational logic. A dedicated buffer receives a target address corresponding to a memory operation instruction. Another dedicated buffer holds a problem address. Yet another dedicated buffer holds a repair address to be used to replace the problem address when memory operations instructions are targeted at that problem address. An address compare combinational circuit compares a received target address held by the dedicated target address buffer with a problem address held by the dedicated repair address buffer. If there is a match, the address compare combinational circuit generates a match notification signal. In response to a match notification signal, an address circuit, which may also be a combinational circuit, responds by causing the repair address within the dedicated repair address buffer to be targeted by the memory operation instruction instead of the problem address being targeted.

Abstract: A memory controller that performs address replacement using combinational logic. A dedicated buffer receives a target address corresponding to a memory operation instruction. Another dedicated buffer holds a problem address. Yet another dedicated buffer holds a repair address to be used to replace the problem address when memory operations instructions are targeted at that problem address. An address compare combinational circuit compares a received target address held by the dedicated target address buffer with a problem address held by the dedicated repair address buffer. If there is a match, the address compare combinational circuit generates a match notification signal. In response to a match notification signal, an address circuit, which may also be a combinational circuit, responds by causing the repair address within the dedicated repair address buffer to be targeted by the memory operation instruction instead of the problem address being targeted.

Abstract: Aspects extend to methods, systems, and computer program products for balancing input phases across server rack power supplies. A rack manager can monitor individual Alternating Current (AC) phase currents at the rack level. The rack manager knows (or can at least determine) which power supplies are connected to which phase. The rack manager can micro adjust individual PSU output voltages to balance current phases at the rack level. Balancing can occur in response to changed server workloads, hot-unplug of one or more servers, etc. When there is one PSU per server, phase balancing can be accomplished by connecting outputs of power supplies together via busbar or wire. Output voltages of individual power supplies can be adjusted to achieve better phase balancing. Phase imbalance can be corrected by a bus bar or wire carrying enough load to correct phase imbalance.

Abstract: Aspects extend to methods, systems, and computer program products for balancing input phases across server rack power supplies. A rack manager can monitor individual Alternating Current (AC) phase currents at the rack level. The rack manager knows (or can at least determine) which power supplies are connected to which phase. The rack manager can micro adjust individual PSU output voltages to balance current phases at the rack level. Balancing can occur in response to changed server workloads, hot-unplug of one or more servers, etc. When there is one PSU per server, phase balancing can be accomplished by connecting outputs of power supplies together via busbar or wire. Output voltages of individual power supplies can be adjusted to achieve better phase balancing. Phase imbalance can be corrected by a bus bar or wire carrying enough load to correct phase imbalance.

Abstract: A data center may use generators for backup power, where the backup generators are able to serve less than the maximal power load of the data center. The use of generators that can serve less than the maximal power load may be recognized by, and addressed by, the power policy for the data center. When utility power is lost, the power policy manager may detect that the data center is switching to generator power, and may have knowledge of the capacity of those generators. If the capacity is less than the maximal power load of the data center, the power policy manager may shed load by throttling equipment in the data center to lower power levels. The policy may specify which pieces of equipment are to be throttled. When utility power is restored, the throttling of the equipment may be removed by the power policy manager.

Abstract: One system may comprise an interleave system that determines a desired interleave for at least a selected portion of a distributed memory system. A migration system is associated with the interleave system to migrate blocks of data from first assigned memory locations of the distributed memory system to second interleaved memory locations of the distributed memory system to implement the desired interleave.

Abstract: An optimized temperature-driven device cooling mechanism factors in power consumption of thermal management components, as well as processing components. A decision engine consults a knowledge base comprising information regarding power consumption and thermal impact of cooling components and processing components. The decision engine can delay additional cooling if the power consumption of cooling components is greater than the increase in power consumption of processing components due to increases in ambient temperature. The knowledge base is populated by published specifications and by empirically derived data that can be based on tests performed on components. Thermal management strategies can be user selected and can include strategies that avoid controlling certain components such as, for example, processing components, instead focusing only on cooling and other components. Cooling includes heat distribution, such as by transferring processing from one processing component to another.

Abstract: The power draw of equipment in a data center may be capped in order to keep the power draw under the capacity of the Uninterruptable Power Supply (UPS) that serves the data center. The current capacity of the UPS may be estimated, and the equipment may be controlled so as to keep the equipment's power draw under that current capacity. Factors that may affect the estimate of the UPS's current capacity include the history of temperature and humidity to which the UPS has been subject, and charge/discharge history of the UPS. Factors that may affect the decision of which equipment to throttle to a lower power level include: the current power load at the data center, the type of software that each server is running, and the demand for that software.

Abstract: A data center may be built from docking stations, and from removable and replaceable IT cartridges. In one example, a docking station is built that provides cooling capability, and other support capability, for IT equipment. An IT cartridge is coupled to the docking station, thereby allowing the IT equipment in the cartridge to receive cooling and other support from the docking station, without the cartridge having any cooling infrastructure (or without having a substantial cooling infrastructure). Since IT equipment may have a shorter useful life than cooling equipment (or other non-IT equipment), when the IT equipment has reached the end of its useful life, the IT cartridge can be retired, and replaced with a new IT cartridge, thereby allowing the longer-lived equipment in the docking station to be easily reused with new IT equipment.

Abstract: A data center may use generators for backup power, where the backup generators are able to serve less than the maximal power load of the data center. The use of generators that can serve less than the maximal power load may be recognized by, and addressed by, the power policy for the data center. When utility power is lost, the power policy manager may detect that the data center is switching to generator power, and may have knowledge of the capacity of those generators. If the capacity is less than the maximal power load of the data center, the power policy manager may shed load by throttling equipment in the data center to lower power levels. The policy may specify which pieces of equipment are to be throttled. When utility power is restored, the throttling of the equipment may be removed by the power policy manager.

Abstract: The power draw of equipment in a data center may be capped in order to keep the power draw under the capacity of the Uninterruptable Power Supply (UPS) that serves the data center. The current capacity of the UPS may be estimated, and the equipment may be controlled so as to keep the equipment's power draw under that current capacity. Factors that may affect the estimate of the UPS's current capacity include the history of temperature and humidity to which the UPS has been subject, and charge/discharge history of the UPS. Factors that may affect the decision of which equipment to throttle to a lower power level include: the current power load at the data center, the type of software that each server is running, and the demand for that software.

Abstract: Embodiments related to a multi-server unit having a direct network topology are disclosed. For example, one disclosed embodiment provides a multi-server unit including a plurality of server nodes connected in a direct network topology including distributed switching between the plurality of server nodes. The plurality of server nodes further comprises a router server node having one or more ports configured to communicate with an outside network, one or more ports configured to communicate with other server nodes of the plurality of server nodes, a logic subsystem, and instructions executable to implement a router configured to direct traffic between the one or more ports configured to communicate with an outside network and the one or more ports configured to communicate with other server nodes of the plurality of server nodes via the direct network.

Abstract: An optimized temperature-driven device cooling mechanism factors in power consumption of thermal management components, as well as processing components. A decision engine consults a knowledge base comprising information regarding power consumption and thermal impact of cooling components and processing components. The decision engine can delay additional cooling if the power consumption of cooling components is greater than the increase in power consumption of processing components due to increases in ambient temperature. The knowledge base is populated by published specifications and by empirically derived data that can be based on tests performed on components. Thermal management strategies can be user selected and can include strategies that avoid controlling certain components such as, for example, processing components, instead focusing only on cooling and other components. Cooling includes heat distribution, such as by transferring processing from one processing component to another.

Abstract: A memory module includes a memory device having a plurality of data pins and conductive lines electrically connected to the plurality of data pins. The memory device is configurable, using at least one input to the memory device, to a data pin configuration selected from among a plurality of different data pin configurations. The plurality of different data pin configurations include a first data pin configuration that uses a first number of data pins of the memory device, and a second data pin configuration that uses a second, different number of data pins.

Abstract: This document describes various techniques for powering a computer data center using a wind-powered generator. The computer data center may include network connected servers that are electrically connected to, and powered by, the wind-powered generator.

Abstract: A scalable computer node includes a first central processing unit (CPU), a memory subsystem, and a socket that is configured to receive a second CPU. An expansion module is mounted in the socket instead of the second CPU, where the expansion module is socket-compatible with the second CPU. The expansion module has a CPU interface to communicate with the first CPU, a memory interface to communicate with the memory subsystem, and a fabric interface to communicate over a communications fabric with an expansion electronic subsystem to expand a capacity of the computer node.

Abstract: A system and method for single hop, processor-to-processor communication in a multiprocessing system over a plurality of crossbars are disclosed. Briefly described, one embodiment is a multiprocessing system comprising a plurality of processors having a plurality of high-bandwidth point-to-point links; a plurality of processor clusters, each processor cluster having a predefined number of the processors residing therein; and a plurality of crossbars, one of the crossbars coupling each of the processors of one of the plurality of processor clusters to each of the processors of another of the plurality of processor clusters, such that all processors are coupled to each of the other processors, and such that the number of crossbars is equal to [X*(X?1)/2], wherein X equals the number of processor clusters.

Abstract: A scalable computer node includes a first central processing unit (CPU), a memory subsystem, and a socket that is configured to receive a second CPU. An expansion module is mounted in the socket instead of the second CPU, where the expansion module is socket-compatible with the second CPU. The expansion module has a CPU interface to communicate with the first CPU, a memory interface to communicate with the memory subsystem, and a fabric interface to communicate over a communications fabric with an expansion electronic subsystem to expand a capacity of the computer node.

Abstract: A memory module includes a memory device having a plurality of data pins and conductive lines electrically connected to the plurality of data pins. The memory device is configurable, using at least one input to the memory device, to a data pin configuration selected from among a plurality of different data pin configurations. The plurality of different data pin configurations include a first data pin configuration that uses a first number of data pins of the memory device, and a second data pin configuration that uses a second, different number of data pins.