A Light-Weight Rollback Mechanism for Testing Kernel Variants in Auto-Tuning

被引:0
|
作者
Hirasawa, Shoichi [1 ,2 ]
Takizawa, Hiroyuki [1 ,2 ]
Kobayashi, Hiroaki [1 ]
机构
[1] Tohoku Univ, Sendai, Miyagi 9808579, Japan
[2] Japan Sci & Technol Agcy, CREST, Tokyo 1020076, Japan
来源
关键词
auto-tuning; rollback; cache; early termination;
D O I
10.1587/transinf.2015PAP0028
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Automatic performance tuning of a practical application could be time-consuming and sometimes infeasible, because it often needs to evaluate the performances of a large number of code variants to find the best one. In this paper, hence, a light-weight rollback mechanism is proposed to evaluate each of code variants at a low cost. In the proposed mechanism, once one code variant of a target code block is executed, the execution state is rolled back to the previous state of not yet executing the block so as to repeatedly execute only the block to find the best code variant. It also has a feature of terminating a code variant whose execution time is longer than the shortest execution time so far. As a result, it can prevent executing the whole application many times and thus reduces the timing overhead of an auto-tuning process required for finding the best code variant.
引用
收藏
页码:2178 / 2186
页数:9
相关论文
共 50 条
  • [1] Testing and Auto-Tuning GPU code with Kernel Tuner
    van Werkhoven, Ben
    2019 18TH INTERNATIONAL SYMPOSIUM ON PARALLEL AND DISTRIBUTED COMPUTING (ISPDC 2019), 2019, : XXI - XXI
  • [2] Auto-tuning Kernel Mean Matching
    Miao, Yun-Qian
    Farahat, Ahmed K.
    Kamel, Mohamed S.
    2013 IEEE 13TH INTERNATIONAL CONFERENCE ON DATA MINING WORKSHOPS (ICDMW), 2013, : 560 - 567
  • [3] Auto-tuning the Matrix Powers Kernel with SEJITS
    Morlan, Jeffrey
    Kamil, Shoaib
    Fox, Armando
    HIGH PERFORMANCE COMPUTING FOR COMPUTATIONAL SCIENCE - VECPAR 2012, 2013, 7851 : 391 - 403
  • [4] Auto-Tuning Structured Light by Optical Stochastic Gradient Descent
    Chen, Wenzheng
    Mirdehghan, Parsa
    Fidler, Sanja
    Kutulakos, Kiriakos N.
    2020 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2020, : 5969 - 5979
  • [5] A novel auto-tuning PID control mechanism for nonlinear systems
    Cetin, Meric
    Iplikci, Serdar
    ISA TRANSACTIONS, 2015, 58 : 292 - 308
  • [6] Multi-Kernel Auto-Tuning on GPUs: Performance and Energy-Aware Optimization
    Guerreiro, Joao
    Ilic, Aleksandar
    Roma, Nuno
    Tomas, Pedro
    23RD EUROMICRO INTERNATIONAL CONFERENCE ON PARALLEL, DISTRIBUTED, AND NETWORK-BASED PROCESSING (PDP 2015), 2015, : 438 - 445
  • [7] Auto-tuning PID/PI control and its application to paper basis weight control
    Tang, Wei
    Shi, Songjiao
    Wang, Mengxiao
    Chung-kuo Tsao Chih/China Pulp and Paper, 2002, 21 (02): : 31 - 35
  • [8] Light-weight resource leak testing based on finalisers
    Dai, Ziying
    Mao, Xiaoguang
    IET SOFTWARE, 2013, 7 (06) : 308 - 316
  • [9] Light-weight kernel instrumentation framework using dynamic binary translation
    Lee, Dongwoo
    Kim, Inhyuk
    Kim, Jeehong
    Jun, Hyung Kook
    Kim, Won Tae
    Lee, Sangwon
    Eom, Young Ik
    JOURNAL OF SUPERCOMPUTING, 2013, 66 (03): : 1613 - 1628
  • [10] Kernel-based MinMax clustering methods with kernelization of the metric and auto-tuning hyper-parameters
    Liu, Junyan
    Guo, Yongan
    Li, Dapeng
    Wang, Zefeng
    Xu, Youyun
    NEUROCOMPUTING, 2019, 359 : 173 - 184