Energy Efficient Task Allocation for Hybrid Main Memory Architecture

Compared with the conventional dynamic random access memory (DRAM), emerging non-volatile memory such as the phase change random access memory (PRAM) has better density and energy efficiency. However, it usually suffers shortcomings of high write power/latency and low write endurance. In this paper, we study the task allocation problem on the hybrid main memory with both DRAM and PRAM in order to leverage the energy consumption and the performance of the memory subsystem. For a static task set, we design an integer linear programming (ILP) based offline allocation algorithm to obtain the optimal task allocation. Furthermore, we propose a heuristic algorithm, which is an offline adaptive space allocation (offline-ASA) algorithm. Experimental results shown that our proposed schemes achieves average 35.1% energy saving, with additional performance cost of 17.1%.