Acceleration of block matching on a low-power heterogeneous multi-core processor based on DTU data-transfer with data re-allocation

Yoshitaka Hiramatsu; Hasitha Muthumala Waidyasooriya; Masanori Hariyama; Toru Nojiri; Kunio Uchiyama; Michitaka Kameyama

doi:10.1587/transele.E95.C.1872

Acceleration of block matching on a low-power heterogeneous multi-core processor based on DTU data-transfer with data re-allocation

Yoshitaka Hiramatsu, Hasitha Muthumala Waidyasooriya, Masanori Hariyama, Toru Nojiri, Kunio Uchiyama, Michitaka Kameyama

INF - Computer and Mathematical Sciences

Research output: Contribution to journal › Article › peer-review

1 Citation (Scopus)

Abstract

The large data-transfer time among different cores is a big problem in heterogeneous multi-core processors. This paper presents a method to accelerate the data transfers exploiting data-transfer-units together with complex memory allocation. We used block matching, which is very common in image processing, to evaluate our technique. The proposed method reduces the data-transfer time by more than 42% compared to the earlier works that use CPU-based data transfers. Moreover, the total processing time is only 15ms for a VGA image with 16 × 16 pixel blocks.

Original language	English
Pages (from-to)	1872-1882
Number of pages	11
Journal	IEICE Transactions on Electronics
Volume	E95-C
Issue number	12
DOIs	https://doi.org/10.1587/transele.E95.C.1872
Publication status	Published - 2012 Dec

Keywords

Accelerator
Block matching
Data transfer
Dynamically reconfigurable processor
Heterogeneous multi-core

ASJC Scopus subject areas

Electronic, Optical and Magnetic Materials
Electrical and Electronic Engineering

Access to Document

10.1587/transele.E95.C.1872

Fingerprint Dive into the research topics of 'Acceleration of block matching on a low-power heterogeneous multi-core processor based on DTU data-transfer with data re-allocation'. Together they form a unique fingerprint.

Cite this

Acceleration of block matching on a low-power heterogeneous multi-core processor based on DTU data-transfer with data re-allocation. / Hiramatsu, Yoshitaka; Waidyasooriya, Hasitha Muthumala ; Hariyama, Masanori; Nojiri, Toru; Uchiyama, Kunio; Kameyama, Michitaka.

In: IEICE Transactions on Electronics, Vol. E95-C, No. 12, 12.2012, p. 1872-1882.

Research output: Contribution to journal › Article › peer-review

@article{99ef7b008fd5432fa2f46b8fbfe9a8ef,

title = "Acceleration of block matching on a low-power heterogeneous multi-core processor based on DTU data-transfer with data re-allocation",

abstract = "The large data-transfer time among different cores is a big problem in heterogeneous multi-core processors. This paper presents a method to accelerate the data transfers exploiting data-transfer-units together with complex memory allocation. We used block matching, which is very common in image processing, to evaluate our technique. The proposed method reduces the data-transfer time by more than 42% compared to the earlier works that use CPU-based data transfers. Moreover, the total processing time is only 15ms for a VGA image with 16 × 16 pixel blocks.",

keywords = "Accelerator, Block matching, Data transfer, Dynamically reconfigurable processor, Heterogeneous multi-core",

author = "Yoshitaka Hiramatsu and Waidyasooriya, {Hasitha Muthumala} and Masanori Hariyama and Toru Nojiri and Kunio Uchiyama and Michitaka Kameyama",

year = "2012",

month = dec,

doi = "10.1587/transele.E95.C.1872",

language = "English",

volume = "E95-C",

pages = "1872--1882",

journal = "IEICE Transactions on Electronics",

issn = "0916-8524",

publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",

number = "12",

}

TY - JOUR

T1 - Acceleration of block matching on a low-power heterogeneous multi-core processor based on DTU data-transfer with data re-allocation

AU - Hiramatsu, Yoshitaka

AU - Waidyasooriya, Hasitha Muthumala

AU - Hariyama, Masanori

AU - Nojiri, Toru

AU - Uchiyama, Kunio

AU - Kameyama, Michitaka

PY - 2012/12

Y1 - 2012/12

N2 - The large data-transfer time among different cores is a big problem in heterogeneous multi-core processors. This paper presents a method to accelerate the data transfers exploiting data-transfer-units together with complex memory allocation. We used block matching, which is very common in image processing, to evaluate our technique. The proposed method reduces the data-transfer time by more than 42% compared to the earlier works that use CPU-based data transfers. Moreover, the total processing time is only 15ms for a VGA image with 16 × 16 pixel blocks.

AB - The large data-transfer time among different cores is a big problem in heterogeneous multi-core processors. This paper presents a method to accelerate the data transfers exploiting data-transfer-units together with complex memory allocation. We used block matching, which is very common in image processing, to evaluate our technique. The proposed method reduces the data-transfer time by more than 42% compared to the earlier works that use CPU-based data transfers. Moreover, the total processing time is only 15ms for a VGA image with 16 × 16 pixel blocks.

KW - Accelerator

KW - Block matching

KW - Data transfer

KW - Dynamically reconfigurable processor

KW - Heterogeneous multi-core

UR - http://www.scopus.com/inward/record.url?scp=84871232925&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871232925&partnerID=8YFLogxK

U2 - 10.1587/transele.E95.C.1872

DO - 10.1587/transele.E95.C.1872

M3 - Article

AN - SCOPUS:84871232925

VL - E95-C

SP - 1872

EP - 1882

JO - IEICE Transactions on Electronics

JF - IEICE Transactions on Electronics

SN - 0916-8524

IS - 12

ER -