Reduction of energy consumption in processors by early detection and bypassing of trivial operations

被引:5
|
作者
Islam, Md. Mafijul [1 ]
Stenstrom, Per [1 ]
机构
[1] Chalmers Univ Technol, Dept Comp Sci & Engn, SE-41296 Gothenburg, Sweden
关键词
D O I
10.1109/ICSAMOS.2006.300805
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Previous research has established that trivial operations, i.e., instructions whose outcome can be trivially inferred from the operands, e.g. addition of zero, account for a quite significant portion of the dynamically executed instructions. By detecting them early and removing them from the pipeline, it is possible to reduce the energy consumption. This paper first presents a new classification of trivial operations in which especially such trivial operations that can be detected early, i.e. at the decode stage, in the pipeline are identified. Our analysis shows that on average as many as 10% of all executed instructions are of this kind across 12 applications from SPEC2000. We find that a majority (indeed 89%) of them are identity-trivial in which at least one of the operands is the identity element - zero or one. By detecting them early, one can bypass their execution and eliminate register accesses if the processor uses a logical/physical register remapping unit. We find that as many as 75% of all trivial operations can be detected and eliminated at the decode stage because the identity element is available that often. With such support, we find that the energy consumption in the functional units, the result bus, the instruction window infrastructure, and the register file can be reduced by 13%, 9%, 27%, and 26%, respectively yielding 18% reduction of the energy in the core pipeline.
引用
收藏
页码:28 / +
页数:2
相关论文
共 50 条
  • [21] Exploring performance and energy consumption differences between recent Intel processors
    Lopez-Novoa, Unai
    2019 IEEE SMARTWORLD, UBIQUITOUS INTELLIGENCE & COMPUTING, ADVANCED & TRUSTED COMPUTING, SCALABLE COMPUTING & COMMUNICATIONS, CLOUD & BIG DATA COMPUTING, INTERNET OF PEOPLE AND SMART CITY INNOVATION (SMARTWORLD/SCALCOM/UIC/ATC/CBDCOM/IOP/SCI 2019), 2019, : 263 - 267
  • [22] Modeling power and energy consumption of dense matrix factorizations on multicore processors
    Alonso, Pedro
    Dolz, Manuel F.
    Mayo, Rafael
    Quintana-Orti, Enrique S.
    CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2014, 26 (17): : 2743 - 2757
  • [23] High-Level Energy Consumption Model of Embedded Graphic Processors
    Huang, Chen-Wei
    Chung, Yu-An
    Huang, Pei-Shu
    Tsao, Shiao-Li
    2015 IEEE INTERNATIONAL CONFERENCE ON DIGITAL SIGNAL PROCESSING (DSP), 2015, : 105 - 109
  • [24] Software energy reduction techniques for variable-voltage processors
    Okuma, T
    Ishihara, T
    Yasuura, H
    IEEE DESIGN & TEST OF COMPUTERS, 2001, 18 (02): : 31 - 41
  • [25] Performance and Power Consumption Investigation for Execution of Integer Operations on CPU and GPU Processors for Multimedia Applications
    Iovanovici, A.
    Visan, C.
    Marcu, M.
    2009 7TH INTERNATIONAL SYMPOSIUM ON INTELLIGENT SYSTEMS AND INFORMATICS, 2009, : 258 - 262
  • [26] Energy consumption reduction of machine tools
    Fujishima, M., 2012, Japan Society for Precision Engineering, Kudan Seiwa Bldg, 1-5-9 Kudan-Kita, Chiyoda-ku, Tokyo, 102-0073, Japan (78):
  • [27] STUDY ON REDUCING ENERGY CONSUMPTION IN ROUGH TURNING OPERATIONS
    Abbasov, I. T.
    Simon, S.
    Fritzsche, P. D.
    Yusubov, N. D.
    SOCAR PROCEEDINGS, 2022, 2022 : 23 - 28
  • [28] Energy consumption model for cutting operations in a stochastic environment
    Mariangela Quarto
    Gianluca D’Urso
    Claudio Giardini
    The International Journal of Advanced Manufacturing Technology, 2020, 110 : 2743 - 2752
  • [29] OPTIMIZING OLEFINS PLANT OPERATIONS TO REDUCE ENERGY CONSUMPTION
    Turk, Martin A.
    2019 PETROLEUM AND CHEMICAL INDUSTRY CONFERENCE EUROPE (PCIC EUROPE), 2019, : 67 - 75
  • [30] Energy consumption model for cutting operations in a stochastic environment
    Quarto, Mariangela
    D'Urso, Gianluca
    Giardini, Claudio
    INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY, 2020, 110 (9-10): : 2743 - 2752