A power routing rack can be used to provide various power support redundancies to one or more computing racks in a data center. Power feeds are supplied to bus bars in a power bus bar array extending through the rack, and power routing assemblies positioned in the rack route power from one or more of the bus bars to a computing rack. Each assembly includes circuit breakers that couple to separate bus bars and a routing module that routes power from one or more of the circuit breakers to one or more computing racks. The routing module can include a transfer switch that selectively routes power. The routing module can include an electrical bridge that concurrently routes power. Each assembly can be positioned within the rack to couple the circuit breakers to various power bus bars to adjust the power support redundancy provided to one or more computing racks.

Embodiments of the present disclosure provide a storage device and a method for powering the storage. The storage device is powered by a main power supply and the storage device comprises a processor, a baseboard management controller (BMC), a plurality of disks, a first backup power supply and a second backup power supply. The first backup power supply is at least coupled to the processor, and the second backup power supply is at least coupled to the baseboard management controller. Embodiments of the present disclosure enable flexible spatial arrangement of backup power supplies by arranging a plurality of backup power supplies in the storage device.

A Green NAND Device (GND) driver application queries AC line and battery status and then stores an image of processor states and caches and a resume routine to DRAM when power failure occurs. A DRAM image is then stored to flash memory for a persistent mode when battery power is available. The image in DRAM may be a partial image that includes entries, flushed caches, processor contexts, ramdisks, write caches, and a resume context. Endurance of flash memory is increased by a Super Enhanced Endurance Device (SEED) SSD. In a power down mode, the GND driver limits DRAM use and only caches in DRAM data that can be deleted on power down. Host accesses to flash are intercepted by the GND driver and categorized by data type. Paging files and temporary files cached in DRAM are optionally written to flash.

An apparatus comprises a write buffer to buffer store requests issued by the processing circuitry, prior to the store data being written to at least one cache. Draining circuitry detects a draining trigger event having potential to cause loss of state stored in the at least one cache. In response to the draining trigger event, the draining circuitry performs a draining operation to identify whether the write buffer buffers any committed store requests requiring persistence, and when the write buffer buffers at least one committed store request requiring persistence, to cause the store data associated with the at least one committed store request to be written to persistent memory. This helps to eliminate barrier instructions from software, simplifying persistent programming and improving performance.

A storage system has a cluster structure in which a node is connected with a different node, the node having a volatile memory for storing first update data from a host and a first non-volatile memory for storing second copy data of second update data from the host to the different node, and having a copy management processing unit for storing first copy data of the first update data into a second non-volatile memory of the different node, and a storage service processing unit for transmitting, to the host, a response with respect to an update request of the first update data in response to the storage of the first copy data of the first update data by the copy management processing unit into the second non-volatile memory of the different node.

A microcontroller system which employs an intermediate approach in hybrid FRAM-SRAM that involves memory mapping of program sections to retain the reliability benefits provided by FRAM while performing almost as efficiently as an SRAM-based system. They system utilizes an energy-aware memory mapping method which maps different program sections to the hybrid FRAM-SRAM MCU such that energy consumption is minimized without sacrificing reliability. The method comprises a memory initialization map, which performs a one-time characterization to find the optimal memory map for the functions that constitute a program. The method further comprises an energy alignment, a hardware/software method that aligns the system's powered-on time intervals to function execution boundaries, which results in further improvements in energy efficiency and performance.

A storage controller has an operating system (OS) and power control firmware configured to manage use of battery power during a power outage event. The OS specifies to the power control firmware first and second sets of physical components that should be shed by power control firmware during a two-phase vault process. Upon a power failure, the power control firmware turns off power to the first set of physical components and notifies the OS of the power failure. The OS determines whether to abort or continue the vault process. If the OS aborts the vault process, the power control firmware restores power to the first set of physical components. If the OS continues the vault process, the power control firmware turns off power to the second set of physical components, the OS saves application state, and moves all data from volatile memory to persistent memory.

Embodiments of systems and methods that ensure integrity of data during unexpected power interruption of loss are disclosed. In some embodiments, critical data is saved quickly and efficiently using backup power. Data integrity is ensured even when the reliability of backup power sources is an issue. In some embodiments, by skipping the updating and saving of system data while operating on backup power, significant reduction of time for saving critical data can be achieved. System data can be restored next time the data storage system is restarted. Improvements of data storage system reliability are thereby attained.

A robot for data logging is described as a module of a portable data transfer system for use in physically transferring very big amounts of data in secure, fast and cheap way. The data logger logs and optionally analyzes sensory and operation data by statistically correlating and combining data, events, and control data from a variety of system modules, user actions, and sensors used to track system transit, handling, operation, and events. The data logger allows forensic analysis and comparison against a mission description to identify system location, transit path, mishandling, tampering, security breaches and problems arising from environmental conditions, design problems, etc. As a result, persons or events causing problems can be identified, retrained, and rectified, and system debugging can solve problems with error in hardware and software. Furthermore, decision and actions can be taken by the robot or by remote control to protect stored data from unauthorized access from third parties, such actions can include destruction of operating keys repository, electrical destruction of storage modules, system shutdown etc.

In one embodiment, a solid state drive (SSD) with power loss protection (PLP) includes a SSD controller, a secondary controller and a power circuit configured to supply power to the SSD from a power source during normal operation and backup power from a backup power source in response to a loss of power supplied by the power source. In the event of a loss of power, the secondary controller is configured to track the holdup time, or duration of time for which the primary controller can operate on backup power. In one embodiment, the holdup time tracked by the secondary controller is stored in a non-volatile memory in communication with the secondary controller.