Read-Tuned STT-RAM and eDRAM Cache Hierarchies for Throughput and Energy Optimization

Abstract

As capacity and complexity of on-chip cache memory hierarchy increases, the service cost to the critical loads from last level cache (LLC), which are frequently repeated, has become a major concern. The processor may stall for a considerable interval while waiting to access the data stored in the cache blocks in LLC, if there are no independent instructions to execute. To provide accelerated service to the critical loads requests from LLC, this paper concentrates on leveraging the additional capacity offered by replacing SRAM-based L2 with spin-transfer torque random access memory (STT-MRAM) to accommodate frequently accessed cache blocks in exclusive read mode in favor of reducing the overall read service time. Our proposed technique improves the temporal locality while preventing cache thrashing via sufficient accommodation of the frequently read reused fraction of working set that may exhibit distant re-reference interval in L2. Our experimental results show that the proposed technique can reduce the L2 read miss ratio by 51.7% on average compared to conventional STT-MRAM L2 design across PARSEC and SPEC2006 workloads while significantly decreasing the L2 dynamic energy consumption.

Keywords

• Non-volatile memory
• STT-MRAM retention relaxation
• last level cache
• energy overhead reduction
• read service time
• critical loads