A Novel High Performance and Energy Efficient NUCA Architecture for STT-MRAM LLCs With Thermal Consideration

被引:6
|
作者
Wu, Bi [1 ,2 ]
Dai, Pengcheng [1 ,2 ]
Cheng, Yuanqing [2 ]
Wang, Ying [3 ]
Yang, Jianlei [1 ,4 ]
Wang, Zhaohao [1 ,2 ]
Liu, Dijun [5 ]
Zhao, Weisheng [1 ,2 ]
机构
[1] Beihang Univ, Fert Beijing Inst, BDBC, Beijing 100191, Peoples R China
[2] Beihang Univ, Sch Microelect, Beijing 100191, Peoples R China
[3] Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China
[4] Beihang Univ, Sch Comp Sci & Engn, Beijing 100191, Peoples R China
[5] China Acad Informat & Commun Technol, Beijing 100191, Peoples R China
基金
中国国家自然科学基金; 北京市自然科学基金;
关键词
System-on-chip; Computer architecture; Magnetic tunneling; Transistors; Switches; Thermal sensors; Organizations; Cache; data migration; low power; spin transfer torque magnetic memory (STT-MRAM); thermal gradient; MAGNETIC TUNNEL-JUNCTION; MEMORY; CACHE; MODEL;
D O I
10.1109/TCAD.2019.2897707
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
As the speed gap of the modern processor and the off-chip main memory enlarges, on-chip cache capacity increases to sustain the performance scaling. As a result, the cache power occupies a large portion of the total power budget. Spin transfer torque magnetic memory (STT-MRAM) is proposed as a promising solution for the low power cache design due to its high integration density and ultralow leakage power. Nevertheless, the high write power and latency of STT-MRAM become new barriers for the commercialization of this emerging technology. In this paper, we investigate the thermal effect on the access performance of STT-MRAM, and observe that the temperature can affect the write delay and energy significantly. Then, we explore the nonuniform cache access (NUCA) design of the chip-multiprocessors with STT-MRAM-based last level cache (LLC). A thermal aware data migration policy, called "Thermosiphon," which takes advantage of the thermal property of STT-MRAM, is proposed to reduce the LLC write energy. This policy splits the LLC into different regions dynamically based on the thermal distribution monitored by thermal sensors available on-chip, and adaptively migrates write intensive data among different thermal regions considering the thermal gradient. Compared to the conventional NUCA design, our proposed design can save 41.2% write energy at most and 13.01% on average with negligible hardware overhead.
引用
收藏
页码:803 / 815
页数:13
相关论文
共 50 条
  • [1] Thermosiphon: A Thermal Aweare NUCA Architecture for Write Energy Reduction of the STT-MRAM based LLCs
    Wu, Bi
    Cheng, Yuanqing
    Dai, Pengcheng
    Yang, Jianlei
    Zhang, Youguang
    Liu, Dijun
    Wang, Ying
    Zhao, Weisheng
    2017 IEEE/ACM INTERNATIONAL CONFERENCE ON COMPUTER-AIDED DESIGN (ICCAD), 2017, : 474 - 481
  • [2] NEAR: A Novel Energy Aware Replacement Policy for STT-MRAM LLCs
    Liu, Chen
    Cheng, Yuanqing
    Wang, Ying
    Zhang, Youguang
    Zhao, Weisheng
    2018 IEEE INTERNATIONAL SYMPOSIUM ON CIRCUITS AND SYSTEMS (ISCAS), 2018,
  • [3] Write Pulse Scaling for Energy Efficient STT-MRAM
    Alkabani, Yousra
    Koopmans, Zach
    Xu, Haifeng
    Jones, Alex K.
    Melhem, Rami
    2016 IEEE COMPUTER SOCIETY ANNUAL SYMPOSIUM ON VLSI (ISVLSI), 2016, : 248 - 253
  • [4] STAxCache: An Approximate, Energy Efficient STT-MRAM Cache
    Ranjan, Ashish
    Venkataramani, Swagath
    Pajouhi, Zoha
    Venkatesan, Rangharajan
    Roy, Kaushik
    Raghunathan, Anand
    PROCEEDINGS OF THE 2017 DESIGN, AUTOMATION & TEST IN EUROPE CONFERENCE & EXHIBITION (DATE), 2017, : 356 - 361
  • [5] An Energy Efficient Multi-Retention STT-MRAM Memory Architecture for IoT Applications
    Jahannia, Belal
    Ghasemi, Seyed Ali
    Farbeh, Hamed
    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS, 2024, 71 (03) : 1431 - 1435
  • [6] Distributed Accumulation based Energy Efficient STT-MRAM based Digital PIM Architecture
    Kim, Dongsu
    Park, Jongsun
    2022 19TH INTERNATIONAL SOC DESIGN CONFERENCE (ISOCC), 2022, : 29 - 30
  • [7] Designing Efficient and High-Performance AI Accelerators With Customized STT-MRAM
    Mishty, Kaniz
    Sadi, Mehdi
    IEEE TRANSACTIONS ON VERY LARGE SCALE INTEGRATION (VLSI) SYSTEMS, 2021, 29 (10) : 1730 - 1742
  • [8] EMACS: Efficient MBIST Architecture for Test and Characterization of STT-MRAM Arrays
    Yoon, Insik
    Chintaluri, Ashwin
    Raychowdhury, Arijit
    PROCEEDINGS 2016 IEEE INTERNATIONAL TEST CONFERENCE (ITC), 2016,
  • [9] Architecture-level energy model for high-capacity STT-MRAM memory
    Lin, Hui
    Yong, Ruoxue
    Jiang, Yanfeng
    INTERNATIONAL JOURNAL OF ELECTRONICS, 2025, 112 (04) : 601 - 616
  • [10] Performance and Energy Optimization on MPSoCs by Enabling STT-MRAM LUTs
    Ding, Hongyuan
    Huang, Miaoqing
    2015 IEEE 23RD ANNUAL INTERNATIONAL SYMPOSIUM ON FIELD-PROGRAMMABLE CUSTOM COMPUTING MACHINES (FCCM), 2015, : 35 - 35