하이브리드 메모리에서의 에너지 수준에 따른 버퍼 할당 알고리즘

Abstract

STT-MRAM (Spin Transfer Torque-Magnetic RAM) has been recently researched to replace DRAM which requires large cell size and high leakage power consumption. Although the read operation in STT-MRAM is acceptable in terms of performance and energy consumption, the write operation discourages the adoption of the STT-MRAM as main memory. One solution for the poor write operation is to reduce retention time of the memory by reducing the cell size. If the memory access pattern in a stream application is considered on a system with multiple retention time STT-MRAMs, the write energy can be reduced. This paper solves the buffer mapping problem onto a system with multiple retention time memories for a stream application to minimize the energy consumption. Experimental results show that a system with two or three different retention time STTMRAMs reduces 30~70% of write energy consumption compared with a single high retention STT-MRAM.