Abstract: A system for achieving memory persistence includes a volatile memory, a non-volatile memory, and a processor. The processor may indicate a volatile memory range for the processor to backup, and open a memory window for the processor to access. The system further includes a power supply. The power supply may provide power for the processor to backup the memory range of the volatile memory. The processor may, responsive to an occurrence of a backup event, initiate a memory transfer using the opened memory window. The memory transfer uses the processor to move the memory range of the volatile memory to a memory region of the non-volatile memory.

Abstract: Example implementations relate to selectively enabling backup power to nodes. For example, a system includes a plurality of compute nodes, where each node includes a plurality of loads. The system also includes at least one backup power supply and control module coupled to the nodes and to the at least one backup power supply. The control module is to selectively enable an output of power from the at least one backup power supply to the nodes based on a comparison of a power capacity of the at least one backup power supply and a backup power demand of the nodes.

Abstract: Techniques for reducing power consumption of a storage controller are provided. An example method includes determining a back-end bandwidth of a storage system, wherein the back-end of the storage system includes a storage drive communicatively coupled to a storage controller. The method also includes determining a front-end bandwidth of the storage system, wherein the front-end of the storage system includes a front-end bus coupling the storage controller to a host. The method also includes computing a target back-end memory bandwidth based on the back-end bandwidth and computing a target front-end memory bandwidth based on the front-end bandwidth. The method also includes reducing power consumption of the storage controller by reducing a clock frequency of a memory device of the storage controller based on the greater of the target back-end memory bandwidth and the target front-end memory bandwidth.

Abstract: A system for achieving memory persistence includes a volatile memory, a non-volatile memory, and a processor. The processor may indicate a volatile memory range for the processor to backup, and open a memory window for the processor to access. The system further includes a power supply. The power supply may provide power for the processor to backup the memory range of the volatile memory. The processor may, responsive to an occurrence of a backup event, initiate a memory transfer using the opened memory window. The memory transfer uses the processor to move the memory range of the volatile memory to a memory region of the non-volatile memory.

Abstract: Techniques to determine an adjustment to front end bandwidth of a server based on backend bandwidth and to adjust power consumption of an input/output (I/O) device.

Abstract: Example implementations relate to selectively enabling backup power to nodes. For example, a system includes a plurality of compute nodes, where each node includes a plurality of loads. The system also includes at least one backup power supply and control module coupled to the nodes and to the at least one backup power supply. The control module is to selectively enable an output of power from the at least one backup power supply to the nodes based on a comparison of a power capacity of the at least one backup power supply and a backup power demand of the nodes.

Abstract: Techniques for reducing power consumption of a storage controller are provided. An example method includes determining a back-end bandwidth of a storage system, wherein the back-end of the storage system includes a storage drive communicatively coupled to a storage controller. The method also includes determining a front-end bandwidth of the storage system, wherein the front-end of the storage system includes a front-end bus coupling the storage controller to a host. The method also includes computing a target back-end memory bandwidth based on the back-end bandwidth and computing a target front-end memory bandwidth based on the front-end bandwidth. The method also includes reducing power consumption of the storage controller by reducing a clock frequency of a memory device of the storage controller based on the greater of the target back-end memory bandwidth and the target front-end memory bandwidth.

Abstract: Techniques to determine an adjustment to front end bandwidth of a server based on backend bandwidth and to adjust power consumption of an input/output (I/O) device.

Abstract: A failure of at least one of a plurality of drives of a redundant array of independent disks (RAID) is determined. A remainder of the plurality of drives of the RAID are remapped. A RAID level of the RAID is maintained, if the remainder of the plurality of drives are remapped.

Abstract: A failure of at least one of a plurality of drives of a redundant array of independent disks (RAID) is determined. A remainder of the plurality of drives of the RAID are remapped. A RAID level of the RAID is maintained, if the remainder of the plurality of drives are remapped.