Energy-Efficient Memory Mappings based on Partial WCET Analysis and Multi-Retention Time STT-RAM - Archive ouverte HAL Access content directly
Conference Papers Year : 2018

Energy-Efficient Memory Mappings based on Partial WCET Analysis and Multi-Retention Time STT-RAM

(1) , (1) , (2)
1
2

Abstract

Energy-efficiency has become one major challenge in both embedded and high-performance computing. Different approaches have been investigated to solve the challenge, e.g., heterogeneous multicore, system runtime and device-level power management. This paper targets emerging non volatile memories (NVMs), through Spin-Transfer Torque RAM (STT-RAM), which inherently have quasi-null leakage. This enables to reduce the static power consumption, which tends to become dominant in modern systems. The usage of NVM in memory hierarchy comes however at the cost of expensive write operations in terms of latency and energy. In order to mitigate this detrimental feature, this paper leverages the notion of delta worst-case execution time (δ-WCET), which consists of partial WCET estimates. From program analysis, δ-WCETs are determined and used to safely allocate data to NVM memory banks with variable data retention times. The δ-WCET analysis computes the WCET between any two locations in a function code, i.e., between basic blocks or instructions. Our approach is validated on the Mälardalen benchmark suite and significant memory dynamic energy reductions (up to 80 %, and 66 % on average) are reported.
Fichier principal
Vignette du fichier
RTNS'18.pdf (1.14 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01871320 , version 1 (10-09-2018)

Identifiers

Cite

Rabab Bouziane, Erven Rohou, Abdoulaye Gamatié. Energy-Efficient Memory Mappings based on Partial WCET Analysis and Multi-Retention Time STT-RAM. RTNS: Real-Time Networks and Systems, Oct 2018, Poitiers, France. pp.148-158, ⟨10.1145/3273905.3273908⟩. ⟨hal-01871320⟩
363 View
245 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More