DPA: Demand-Based Partition and Data Allocation for Hybrid On-Chip Memory

Jianfeng Wang,Leibo Liu,Jianfeng Zhu,Shaojun Wei
DOI: https://doi.org/10.1109/ccet50901.2020.9213108
2020-01-01
Abstract:The performance and energy consumption of the memory system has become an essential bottleneck in modern processors. Scratchpad memory (SPM) is widely used in the embedded system for meeting the constrict of energy consumption. The hybrid memory architecture has been proposed to improve the energy efficiency of the overall system. Previous works focus on optimizing data allocations in the fixed-size cache and SPM. However, applications have different preferences for cache and SPM. Fixed memory structures may cause a waste of storage resources. In this paper, a reconfigurable mechanism is proposed to dynamically adjust the partition of the local memory and allocate data with high-locality into the SPM. Our experimental results show that our method improves the energy-delay product by 12.0% on average compared with the previous works.
What problem does this paper attempt to address?