Presentation 2008-05-13
Speculation scheme that continues executing mispredicted instructions
Takanobu KITA, Ryota SHIOYA, Eiji IRIE, Masahiro GOSHIMA, Shuichi SAKAI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Modern processors are acquiring deeper pipelines as their clock frequencies grow higher, leading to large misprediction penalties. Not all instructions flushed on misprediction, however, are wrongly executed ; there are instructions independent of control flow. Not flushing such instructions and leaving them for continued execution will reduce misprediction penalty. This paper proposes a speculation scheme which continues execution of instructions on misprediction instead of flushing them. The results of instructions independent of control flow may be used for branch prediction and value prediction, and help execution of correct path. The evaluation showed that with an addition of small amount of Hardware, the proposed scheme will increase IPCs by average of 3.8% (12SPECint2000, SPECint2006 and MediaBench.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) branch prediction / misprediction
Paper # ICD2008-18
Date of Issue

Conference Information
Committee ICD
Conference Date 2008/5/6(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Integrated Circuits and Devices (ICD)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Speculation scheme that continues executing mispredicted instructions
Sub Title (in English)
Keyword(1) branch prediction
Keyword(2) misprediction
1st Author's Name Takanobu KITA
1st Author's Affiliation S. Sakai and M. Goshima Lab., Department of Information and Communication Engineering, Graduate School of Information Science and Technology, The University of Tokyo()
2nd Author's Name Ryota SHIOYA
2nd Author's Affiliation S. Sakai and M. Goshima Lab., Department of Information and Communication Engineering, Graduate School of Information Science and Technology, The University of Tokyo
3rd Author's Name Eiji IRIE
3rd Author's Affiliation S. Sakai and M. Goshima Lab., Department of Information and Communication Engineering, Graduate School of Information Science and Technology, The University of Tokyo
4th Author's Name Masahiro GOSHIMA
4th Author's Affiliation S. Sakai and M. Goshima Lab., Department of Information and Communication Engineering, Graduate School of Information Science and Technology, The University of Tokyo
5th Author's Name Shuichi SAKAI
5th Author's Affiliation S. Sakai and M. Goshima Lab., Department of Information and Communication Engineering, Graduate School of Information Science and Technology, The University of Tokyo
Date 2008-05-13
Paper # ICD2008-18
Volume (vol) vol.108
Number (no) 28
Page pp.pp.-
#Pages 6
Date of Issue