GPU LOW-POWER GPU SCHEDULING METHOD FOR DISTRIBUTED STORAGE APPLICATION

摘要

Provided is a low power GPU scheduling method for a distributed storage application. A low power GPU scheduling method for a distributed storage application according to an embodiment includes the steps of: generating, by an encoding scheduler of an initiator server, a parity block from a data block stored in a CPU memory by calling a GPU parity calculation function; switching, by the encoding scheduler of the initiator server, a power mode of an HDD from a standby mode to an active mode in order to strip the parity block on an HDD storage; striping, by the encoding scheduler of the initiator server, the parity blocks on a parity HDD; and switching, by the encoding scheduler of the initiator server, the power mode of the HDD from the active mode to the standby mode to reduce power consumption.