Issue mechanism for embedded simultaneous multithreading processor

Chengjie Zang, Shigeki Imai, Steven Frank, Shinji Kimura

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Simultaneous Multithreading (SMT) technology enhances instruction throughput by issuing multiple instructions from multiple threads within one clock cycle. For in-order pipeline to each thread, SMT processors can providelarge number of issued instructions close to or surpass than using out-of-order pipeline. In this work, we show an efficient issue logic for predicated instruction sequence with the parallel flag in each instruction, where the predicate register based issue control is adopted and the continuous instructions with the parallel flag of '0' are executed in parallel. The flag is pre-defined by a compiler. Instructions from different threads are issuedbased on the round-robin order. We also introduce an Instruction Queue skipmechanism for thread if the queue is empty. Using this kind of issue logic,we designed a 6 threads, 7-stage, in-order pipeline processor. Based on this processor, we compare round-robin issue policy (RR(T1-Tn)) with other policies: thread one always has the highest priority (PR(T1)) and thread one or thread n has the highest priority in turn (PR(T1-T n)). The results show that RR(T1-Tn) policy outperforms others and PR(T1-Tn) is almost the same to RR(T1-Tn) from the point of view of the issued instructions per cycle.

Original languageEnglish
Pages (from-to)1092-1100
Number of pages9
JournalIEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
VolumeE91-A
Issue number4
DOIs
Publication statusPublished - 2008

Fingerprint

Multithreading
Thread
Pipelines
Clocks
Throughput
Queue
Logic
Cycle
Compiler
Predicate

Keywords

  • Balance roundrobin policy
  • Parallel flag
  • Simultaneous multithreading

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Graphics and Computer-Aided Design
  • Applied Mathematics
  • Signal Processing

Cite this

Issue mechanism for embedded simultaneous multithreading processor. / Zang, Chengjie; Imai, Shigeki; Frank, Steven; Kimura, Shinji.

In: IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, Vol. E91-A, No. 4, 2008, p. 1092-1100.

Research output: Contribution to journalArticle

@article{bb26a9ae2c9d43e5a75f9a2f9f0d5153,
title = "Issue mechanism for embedded simultaneous multithreading processor",
abstract = "Simultaneous Multithreading (SMT) technology enhances instruction throughput by issuing multiple instructions from multiple threads within one clock cycle. For in-order pipeline to each thread, SMT processors can providelarge number of issued instructions close to or surpass than using out-of-order pipeline. In this work, we show an efficient issue logic for predicated instruction sequence with the parallel flag in each instruction, where the predicate register based issue control is adopted and the continuous instructions with the parallel flag of '0' are executed in parallel. The flag is pre-defined by a compiler. Instructions from different threads are issuedbased on the round-robin order. We also introduce an Instruction Queue skipmechanism for thread if the queue is empty. Using this kind of issue logic,we designed a 6 threads, 7-stage, in-order pipeline processor. Based on this processor, we compare round-robin issue policy (RR(T1-Tn)) with other policies: thread one always has the highest priority (PR(T1)) and thread one or thread n has the highest priority in turn (PR(T1-T n)). The results show that RR(T1-Tn) policy outperforms others and PR(T1-Tn) is almost the same to RR(T1-Tn) from the point of view of the issued instructions per cycle.",
keywords = "Balance roundrobin policy, Parallel flag, Simultaneous multithreading",
author = "Chengjie Zang and Shigeki Imai and Steven Frank and Shinji Kimura",
year = "2008",
doi = "10.1093/ietfec/e91-a.4.1092",
language = "English",
volume = "E91-A",
pages = "1092--1100",
journal = "IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences",
issn = "0916-8508",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "4",

}

TY - JOUR

T1 - Issue mechanism for embedded simultaneous multithreading processor

AU - Zang, Chengjie

AU - Imai, Shigeki

AU - Frank, Steven

AU - Kimura, Shinji

PY - 2008

Y1 - 2008

N2 - Simultaneous Multithreading (SMT) technology enhances instruction throughput by issuing multiple instructions from multiple threads within one clock cycle. For in-order pipeline to each thread, SMT processors can providelarge number of issued instructions close to or surpass than using out-of-order pipeline. In this work, we show an efficient issue logic for predicated instruction sequence with the parallel flag in each instruction, where the predicate register based issue control is adopted and the continuous instructions with the parallel flag of '0' are executed in parallel. The flag is pre-defined by a compiler. Instructions from different threads are issuedbased on the round-robin order. We also introduce an Instruction Queue skipmechanism for thread if the queue is empty. Using this kind of issue logic,we designed a 6 threads, 7-stage, in-order pipeline processor. Based on this processor, we compare round-robin issue policy (RR(T1-Tn)) with other policies: thread one always has the highest priority (PR(T1)) and thread one or thread n has the highest priority in turn (PR(T1-T n)). The results show that RR(T1-Tn) policy outperforms others and PR(T1-Tn) is almost the same to RR(T1-Tn) from the point of view of the issued instructions per cycle.

AB - Simultaneous Multithreading (SMT) technology enhances instruction throughput by issuing multiple instructions from multiple threads within one clock cycle. For in-order pipeline to each thread, SMT processors can providelarge number of issued instructions close to or surpass than using out-of-order pipeline. In this work, we show an efficient issue logic for predicated instruction sequence with the parallel flag in each instruction, where the predicate register based issue control is adopted and the continuous instructions with the parallel flag of '0' are executed in parallel. The flag is pre-defined by a compiler. Instructions from different threads are issuedbased on the round-robin order. We also introduce an Instruction Queue skipmechanism for thread if the queue is empty. Using this kind of issue logic,we designed a 6 threads, 7-stage, in-order pipeline processor. Based on this processor, we compare round-robin issue policy (RR(T1-Tn)) with other policies: thread one always has the highest priority (PR(T1)) and thread one or thread n has the highest priority in turn (PR(T1-T n)). The results show that RR(T1-Tn) policy outperforms others and PR(T1-Tn) is almost the same to RR(T1-Tn) from the point of view of the issued instructions per cycle.

KW - Balance roundrobin policy

KW - Parallel flag

KW - Simultaneous multithreading

UR - http://www.scopus.com/inward/record.url?scp=78049330656&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78049330656&partnerID=8YFLogxK

U2 - 10.1093/ietfec/e91-a.4.1092

DO - 10.1093/ietfec/e91-a.4.1092

M3 - Article

VL - E91-A

SP - 1092

EP - 1100

JO - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

JF - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

SN - 0916-8508

IS - 4

ER -