Optimizing Data Allocation and Memory Configuration for Non-Volatile Memory Based Hybrid SPM on Embedded CMPs

The recent emergence of various Non-Volatile Memories (NVMs), with many attractive characteristics such as low leakage power and high-density, provides us with a new way of addressing the memory power consumption problem. In this paper, we target embedded CMPs, and propose a novel Hybrid Scratch Pad Memory (HSPM) architecture which consists of SRAM and NVM to take advantage of the ultra-low leakage power, high density of NVM and fast read of SRAM. A novel data allocation algorithm as well as an algorithm to determine NVM/SRAM ratio for the novel HSPM architecture are proposed. The experimental results show that the data allocation algorithm can reduce the memory access time by 31.22% and the dynamic energy consumption by 15.35% on average for the HSPM architecture when compared with a greedy algorithm.