A trace cache with DVFS techniques for a low power microprocessor

Hyung Beom Jang, Lynn Choi, Sung Woo Jung

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

The trace cache is a solution to achieving high instruction fetches bandwidth by buffering and reusing dynamic instruction traces. This work presents a new trace cache implementation that includes the DVFS (Dynamic Voltage and Frequency Scaling) techniques for energy efficiency. The focus of this paper is to compare the trace cache with DVFS techniques to the conventional trace cache organization where any DVFS technique is not applied. Instead of storing the basic blocks in the unified trace cache space, the first block of each trace is stored in the specific space of the trace cache and the other basic blocks are stored in the rest of the trace cache space. The first basic block area is not voltage scaled because the first basic block should be supplied to processor's front-end as soon as possible. On the other hand, other basic block area is voltage-scaled down in order to reduce the power consumption. Transistor switching speed of other basic block area is slower than that of the first basic block area due to the lowered supply voltage. Our experiments show that when we adopted the DVFS techniques to the conventional trace cache, 12.8% fetch engine energy consumption is reduced, on average. Applying different supply voltages to each different region of the trace cache, we can reduce the dynamic power consumption. However, we can know that the region which is supplied with lowered voltage inevitably deteriorates the performance of the trace cache by 5.7%.

Original languageEnglish
Title of host publicationProceedings - 3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008
Pages587-592
Number of pages6
Volume1
DOIs
Publication statusPublished - 2008 Dec 29
Event3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008 - Busan, Korea, Republic of
Duration: 2008 Nov 112008 Nov 13

Other

Other3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008
CountryKorea, Republic of
CityBusan
Period08/11/1108/11/13

Fingerprint

Microprocessor chips
Electric potential
Electric power utilization
Energy efficiency
Transistors
Energy utilization
Voltage scaling
Dynamic frequency scaling
Engines
Bandwidth
Experiments

ASJC Scopus subject areas

  • Information Systems
  • Software

Cite this

Jang, H. B., Choi, L., & Jung, S. W. (2008). A trace cache with DVFS techniques for a low power microprocessor. In Proceedings - 3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008 (Vol. 1, pp. 587-592). [4682090] https://doi.org/10.1109/ICCIT.2008.148

A trace cache with DVFS techniques for a low power microprocessor. / Jang, Hyung Beom; Choi, Lynn; Jung, Sung Woo.

Proceedings - 3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008. Vol. 1 2008. p. 587-592 4682090.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Jang, HB, Choi, L & Jung, SW 2008, A trace cache with DVFS techniques for a low power microprocessor. in Proceedings - 3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008. vol. 1, 4682090, pp. 587-592, 3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008, Busan, Korea, Republic of, 08/11/11. https://doi.org/10.1109/ICCIT.2008.148
Jang HB, Choi L, Jung SW. A trace cache with DVFS techniques for a low power microprocessor. In Proceedings - 3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008. Vol. 1. 2008. p. 587-592. 4682090 https://doi.org/10.1109/ICCIT.2008.148
Jang, Hyung Beom ; Choi, Lynn ; Jung, Sung Woo. / A trace cache with DVFS techniques for a low power microprocessor. Proceedings - 3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008. Vol. 1 2008. pp. 587-592
@inproceedings{0e47063b23674fadb92e9010c06e409d,
title = "A trace cache with DVFS techniques for a low power microprocessor",
abstract = "The trace cache is a solution to achieving high instruction fetches bandwidth by buffering and reusing dynamic instruction traces. This work presents a new trace cache implementation that includes the DVFS (Dynamic Voltage and Frequency Scaling) techniques for energy efficiency. The focus of this paper is to compare the trace cache with DVFS techniques to the conventional trace cache organization where any DVFS technique is not applied. Instead of storing the basic blocks in the unified trace cache space, the first block of each trace is stored in the specific space of the trace cache and the other basic blocks are stored in the rest of the trace cache space. The first basic block area is not voltage scaled because the first basic block should be supplied to processor's front-end as soon as possible. On the other hand, other basic block area is voltage-scaled down in order to reduce the power consumption. Transistor switching speed of other basic block area is slower than that of the first basic block area due to the lowered supply voltage. Our experiments show that when we adopted the DVFS techniques to the conventional trace cache, 12.8{\%} fetch engine energy consumption is reduced, on average. Applying different supply voltages to each different region of the trace cache, we can reduce the dynamic power consumption. However, we can know that the region which is supplied with lowered voltage inevitably deteriorates the performance of the trace cache by 5.7{\%}.",
author = "Jang, {Hyung Beom} and Lynn Choi and Jung, {Sung Woo}",
year = "2008",
month = "12",
day = "29",
doi = "10.1109/ICCIT.2008.148",
language = "English",
isbn = "9780769534077",
volume = "1",
pages = "587--592",
booktitle = "Proceedings - 3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008",

}

TY - GEN

T1 - A trace cache with DVFS techniques for a low power microprocessor

AU - Jang, Hyung Beom

AU - Choi, Lynn

AU - Jung, Sung Woo

PY - 2008/12/29

Y1 - 2008/12/29

N2 - The trace cache is a solution to achieving high instruction fetches bandwidth by buffering and reusing dynamic instruction traces. This work presents a new trace cache implementation that includes the DVFS (Dynamic Voltage and Frequency Scaling) techniques for energy efficiency. The focus of this paper is to compare the trace cache with DVFS techniques to the conventional trace cache organization where any DVFS technique is not applied. Instead of storing the basic blocks in the unified trace cache space, the first block of each trace is stored in the specific space of the trace cache and the other basic blocks are stored in the rest of the trace cache space. The first basic block area is not voltage scaled because the first basic block should be supplied to processor's front-end as soon as possible. On the other hand, other basic block area is voltage-scaled down in order to reduce the power consumption. Transistor switching speed of other basic block area is slower than that of the first basic block area due to the lowered supply voltage. Our experiments show that when we adopted the DVFS techniques to the conventional trace cache, 12.8% fetch engine energy consumption is reduced, on average. Applying different supply voltages to each different region of the trace cache, we can reduce the dynamic power consumption. However, we can know that the region which is supplied with lowered voltage inevitably deteriorates the performance of the trace cache by 5.7%.

AB - The trace cache is a solution to achieving high instruction fetches bandwidth by buffering and reusing dynamic instruction traces. This work presents a new trace cache implementation that includes the DVFS (Dynamic Voltage and Frequency Scaling) techniques for energy efficiency. The focus of this paper is to compare the trace cache with DVFS techniques to the conventional trace cache organization where any DVFS technique is not applied. Instead of storing the basic blocks in the unified trace cache space, the first block of each trace is stored in the specific space of the trace cache and the other basic blocks are stored in the rest of the trace cache space. The first basic block area is not voltage scaled because the first basic block should be supplied to processor's front-end as soon as possible. On the other hand, other basic block area is voltage-scaled down in order to reduce the power consumption. Transistor switching speed of other basic block area is slower than that of the first basic block area due to the lowered supply voltage. Our experiments show that when we adopted the DVFS techniques to the conventional trace cache, 12.8% fetch engine energy consumption is reduced, on average. Applying different supply voltages to each different region of the trace cache, we can reduce the dynamic power consumption. However, we can know that the region which is supplied with lowered voltage inevitably deteriorates the performance of the trace cache by 5.7%.

UR - http://www.scopus.com/inward/record.url?scp=57849096210&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=57849096210&partnerID=8YFLogxK

U2 - 10.1109/ICCIT.2008.148

DO - 10.1109/ICCIT.2008.148

M3 - Conference contribution

SN - 9780769534077

VL - 1

SP - 587

EP - 592

BT - Proceedings - 3rd International Conference on Convergence and Hybrid Information Technology, ICCIT 2008

ER -