Abstract: A power supply unit (PSU) dynamically limits total recovery current. The PSU includes at least a power input, a power output, a historic maximum power draw memory, an update logic, and a recovery current limiting logic. Some implementations include a latest power measurement register, an hourly max power register, and a rolling max register, and controlling firmware. The update logic monitors a power level. The update logic updates the historic maximum power draw memory to match the monitored level. After a power interruption, the recovery current permitted to flow into the PSU is limited based on the historic usage. The recovery current may be limited in a constant, stepped, or ramped manner. The PSU may also provide power distribution. Multiple PSUs may be treated as a group, allowing an individual PSU to exceed its historic usage while the group's recovery currents are limited to the sum of historic usage levels.

Abstract: A power supply is described herein which provides power to a load, such as a load including one or more computing devices. The power supply uses a slow-response power source (such as a fuel-driven mechanism) to handle a slow-moving component of the demand level presented by the load, and uses a fast-response power source (such as a battery or a capacitor, etc.) to handle a fast-moving component of the demand level. By virtue of this approach, the power supply can manage the load level as it appears to the slow-response power source, allowing, in turn, the slow-response power source to service even fast-changing loads—a task which it could not otherwise perform due to its native limitations.

Abstract: A method of managing a power supply system for a data center includes circulating a fluid in a cooling circuit, obtaining data regarding a server located in the data center using a sensor, controlling the transfer of heat energy from the server to the fluid based on the data, coupling the fluid to an electrochemical power generator, and generating power for the server using the fluid in the electrochemical power generator.

Abstract: A power supply system for a data center includes a cooling circuit, an electrochemical power generator, a sensor, and a processor. The cooling circuit includes a fluid configured to receive heat energy generated by a server located in the data center. The electrochemical power generator is configured to receive and/or generate the fluid of the cooling circuit and to generate electrical energy for the server using the fluid. The sensor is configured to obtain data regarding the server. The processor is configured to control an amount of heat energy transferred from the server to the fluid based on the data.

Abstract: A power supply system for a data center includes a cooling circuit, an electrochemical power generator, a sensor, and a processor. The cooling circuit includes a fluid configured to receive heat energy generated by a server located in the data center. The electrochemical power generator is configured to receive and/or generate the fluid of the cooling circuit and to generate electrical energy for the server using the fluid. The sensor is configured to obtain data regarding the server. The processor is configured to control an amount of heat energy transferred from the server to the fluid based on the data.

Abstract: Aspects extend to methods, systems, and computer program products for remediating power loss at a server. Aspects of the invention increase the likelihood of gracefully shutting down a server and associated components in a data center when mains power is lost for a specified amount of time (e.g., an amount of time beyond transition to generator power). A server can include a management module (e.g., a BMC) and a watchdog module. When the management controller detects loss of power at a power supply unit, the management controller orchestrates a graceful shutdown of the server in response to power loss. When the management module is unresponsive, the watchdog module provides backup functionality for orchestrating a graceful shutdown in response to power loss. As such, data can be saved from RAM to more durable storage even when the management module is unresponsive.

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 power supply system for a data center includes a cooling circuit, an electrochemical power generator, a sensor, and a processor. The cooling circuit includes a fluid configured to receive heat energy generated by a server located in the data center. The electrochemical power generator is configured to receive and/or generate the fluid of the cooling circuit and to generate electrical energy for the server using the fluid. The sensor is configured to obtain data regarding the server. The processor is configured to control an amount of heat energy transferred from the server to the fluid based on the data.

Abstract: Computing devices receive power from multiple fuel cells, consuming natural gas and outputting electrical energy natively consumable by the computing devices. The fuel cells are sized to provide power to a set of computing devices, such as a rack thereof. The computing devices of a failed fuel cell can receive power from adjacent fuel cells. Additionally, the fuel cells and computing devices are positioned to realize thermal symbiotic efficiencies. Controllers instruct the computing devices to deactivate or throttle down power consuming functions during instances where the power consumption demand is increasing faster than the power being sourced by fuel cells, and instruct the computing devices to activate or throttle up power consuming functions during instances where the power consumption demand is decreasing faster than the power being sourced by the fuel cells. Supplemental power sources, supplementing the fuel cells' inability to quickly change power output, are not required.

Abstract: Computing devices receive power from multiple fuel cells, consuming natural gas and outputting electrical energy natively consumable by the computing devices. The fuel cells are sized to provide power to a set of computing devices, such as a rack thereof. The computing devices of a failed fuel cell can receive power from adjacent fuel cells. Additionally, the fuel cells and computing devices are positioned to realize thermal symbiotic efficiencies. Controllers instruct the computing devices to deactivate or throttle down power consuming functions during instances where the power consumption demand is increasing faster than the power being sourced by fuel cells, and instruct the computing devices to activate or throttle up power consuming functions during instances where the power consumption demand is decreasing faster than the power being sourced by the fuel cells. Supplemental power sources, supplementing the fuel cells' inability to quickly change power output, are not required.

Abstract: Described is a technology by which magnetic flux is used to provide backup power. A transformer has a line power source controllably coupled to a first input winding, and secondary power source controllably coupled to a second input winding. A controller monitors the line power and switches to the secondary power source if the line power voltage drops too low, or uses the secondary power source to augment the line power source if the line power current gets too high. Also described is incrementally transitioning from the secondary power source back to the line power source.

Abstract: Described is a technology by which magnetic flux is used to provide backup power. A transformer has a line power source controllably coupled to a first input winding, and secondary power source controllably coupled to a second input winding. A controller monitors the line power and switches to the secondary power source if the line power voltage drops too low, or uses the secondary power source to augment the line power source if the line power current gets too high. Also described is incrementally transitioning from the secondary power source back to the line power source.

Abstract: A method described herein includes an act of receiving data that is indicative of predicted weather conditions for a particular geographic region, wherein the particular geographic region has an energy generation system therein, and wherein the energy generation system utilizes at least one renewable energy resource to generate electrical power. The method also includes the act of scheduling a computational workload for at least one computer in a data center based at least in part upon the data that is indicative of the predicted weather conditions for the particular geographic region.

Abstract: A power supply system for a data center includes a cooling circuit, an electrochemical power generator, a sensor, and a processor. The cooling circuit includes a fluid configured to receive heat energy generated by a server located in the data center. The electrochemical power generator is configured to receive and/or generate the fluid of the cooling circuit and to generate electrical energy for the server using the fluid. The sensor is configured to obtain data regarding the server. The processor is configured to control an amount of heat energy transferred from the server to the fluid based on the data.

Abstract: A method described herein includes an act of receiving data that is indicative of predicted weather conditions for a particular geographic region, wherein the particular geographic region has an energy generation system therein, and wherein the energy generation system utilizes at least one renewable energy resource to generate electrical power. The method also includes the act of scheduling a computational workload for at least one computer in a data center based at least in part upon the data that is indicative of the predicted weather conditions for the particular geographic region.

Abstract: Computational enclosures may be designed to distribute power from power supplies to load units (e.g., processors, storage devices, or network routers). The architecture may affect the efficiency, cost, modularity, accessibility, and space utilization of the components within the enclosure. Presented herein are power distribution architectures involving a distribution board oriented along a first (e.g., vertical) axis within the enclosure, comprising a power interconnect configured to distribute power among a set of load boards oriented along a second (e.g., lateral) axis and respectively connecting with a set of load units oriented along a third (e.g., sagittal) axis, and a set of power supplies also oriented along the third axis. This orientation may compactly and proximately position the loads near the power supplies in the distribution system, and result in a comparatively low local current that enables the use of printed circuit boards for the distribution board and load boards.

Abstract: Described herein are various technologies pertaining to predicting an amount of electrical power that is to be generated by a power system at a future point in time, wherein the power system utilizes a renewable energy resource to generate electrical power. A camera is positioned to capture an image of sky over a geographic region of interest. The image is analyzed to predict an amount of solar radiation that is to be received by the power source at a future point in time. The predicted solar radiation is used to predict an amount of electrical power that will be output by the power system at the future point in time. A computational resource of a data center that is powered by way of the power source is managed as a function of the predicted amount of power.

Abstract: Various technologies described herein pertain to racking equipment in a data center. A modular equipment rack system can include an upper track, a lower track, a vertical support, a power and network distribution unit, and a tray. The upper track and the lower track can respectively include incrementally spaced mounting locations at which the vertical support and the power and network distribution unit can be attachable. The tray can be attachable to the vertical support and the power and network distribution unit when the vertical support is attached to the upper track at a first upper mounting location and attached to the lower track at a corresponding first lower mounting location, and the power and network distribution unit is attached to the upper track at a second upper mounting location and attached to the lower track at a corresponding second lower mounting location.

Abstract: A power supply is described herein which provides power to a load, such as a load including one or more computing devices. The power supply uses a slow-response power source (such as a fuel-driven mechanism) to handle a slow-moving component of the demand level presented by the load, and uses a fast-response power source (such as a battery or a capacitor, etc.) to handle a fast-moving component of the demand level. By virtue of this approach, the power supply can manage the load level as it appears to the slow-response power source, allowing, in turn, the slow-response power source to service even fast-changing loads—a task which it could not otherwise perform due to its native limitations.

Abstract: A framework that enables a local computing cloud infrastructure for rural (and third world) populations with the ability to connect into the global cloud. The framework include is a low cost architecture of long distance, wireless based, renewable energy powered, and small datacenter (DC) (referred to as a pico-DC) nodes that can fully operate off-grid, both power-wise and Internet connection-wise at a very low cost. Additionally, the framework includes power management and storage techniques that effectively enable low power and efficient power use. Thus, systems are self-sufficient, low maintenance and weather proof with no need for power or data connections.