Time-multiplexed execution on the dynamically reconfigurable processor - A performance/cost evaluation

Yohei Hasegawa; Shohei Abe; Katsuaki Deguchi; Masayasu Suzuki; Hideharu Amano

Time-multiplexed execution on the dynamically reconfigurable processor - A performance/cost evaluation

Yohei Hasegawa, Shohei Abe, Katsuaki Deguchi, Masayasu Suzuki, Hideharu Amano

情報工学科

研究成果: Paper › 査読

抄録

Dynamically Reconfigurable Processor (DRP) developed by NEC Electronics is a coarse grain reconfigurable processor that selects a data path from the on-chip repository of sixteen circuit configurations, or contexts, to implement different logic on one single DRP chip. The impact of time-multiplexed execution to performance and cost is analyzed based on real designs including an IPsec router. The Parallelism Diagram which shows the required PEs in each step of the algorithm is introduced as the basis of the analysis, and models for performance and cost are shown. Evaluation results show that the time-multiplexed execution improves the performance per cost around 4.5 to 14 times than that of the case without time-multiplexed execution.

本文言語	English
ページ数	1
出版ステータス	Published - 2005 6月 20
イベント	ACM/SIGDA Thirteenth ACM International Symposium on Field Programmable Gate Arrays - FPGA 2005 - Monterey, CA, United States 継続期間: 2005 2月 20 → 2005 2月 22

Conference

Conference	ACM/SIGDA Thirteenth ACM International Symposium on Field Programmable Gate Arrays - FPGA 2005
国/地域	United States
City	Monterey, CA
Period	05/2/20 → 05/2/22

ASJC Scopus subject areas

コンピュータサイエンス（全般）

他のファイルとリンク

引用スタイル

@conference{7fe2bf9ac8ed46f980596ff695cacabc,

title = "Time-multiplexed execution on the dynamically reconfigurable processor - A performance/cost evaluation",

abstract = "Dynamically Reconfigurable Processor (DRP) developed by NEC Electronics is a coarse grain reconfigurable processor that selects a data path from the on-chip repository of sixteen circuit configurations, or contexts, to implement different logic on one single DRP chip. The impact of time-multiplexed execution to performance and cost is analyzed based on real designs including an IPsec router. The Parallelism Diagram which shows the required PEs in each step of the algorithm is introduced as the basis of the analysis, and models for performance and cost are shown. Evaluation results show that the time-multiplexed execution improves the performance per cost around 4.5 to 14 times than that of the case without time-multiplexed execution.",

author = "Yohei Hasegawa and Shohei Abe and Katsuaki Deguchi and Masayasu Suzuki and Hideharu Amano",

year = "2005",

month = jun,

day = "20",

language = "English",

note = "ACM/SIGDA Thirteenth ACM International Symposium on Field Programmable Gate Arrays - FPGA 2005 ; Conference date: 20-02-2005 Through 22-02-2005",

}

TY - CONF

T1 - Time-multiplexed execution on the dynamically reconfigurable processor - A performance/cost evaluation

AU - Hasegawa, Yohei

AU - Abe, Shohei

AU - Deguchi, Katsuaki

AU - Suzuki, Masayasu

AU - Amano, Hideharu

PY - 2005/6/20

Y1 - 2005/6/20

N2 - Dynamically Reconfigurable Processor (DRP) developed by NEC Electronics is a coarse grain reconfigurable processor that selects a data path from the on-chip repository of sixteen circuit configurations, or contexts, to implement different logic on one single DRP chip. The impact of time-multiplexed execution to performance and cost is analyzed based on real designs including an IPsec router. The Parallelism Diagram which shows the required PEs in each step of the algorithm is introduced as the basis of the analysis, and models for performance and cost are shown. Evaluation results show that the time-multiplexed execution improves the performance per cost around 4.5 to 14 times than that of the case without time-multiplexed execution.

AB - Dynamically Reconfigurable Processor (DRP) developed by NEC Electronics is a coarse grain reconfigurable processor that selects a data path from the on-chip repository of sixteen circuit configurations, or contexts, to implement different logic on one single DRP chip. The impact of time-multiplexed execution to performance and cost is analyzed based on real designs including an IPsec router. The Parallelism Diagram which shows the required PEs in each step of the algorithm is introduced as the basis of the analysis, and models for performance and cost are shown. Evaluation results show that the time-multiplexed execution improves the performance per cost around 4.5 to 14 times than that of the case without time-multiplexed execution.

UR - http://www.scopus.com/inward/record.url?scp=20344393406&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=20344393406&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:20344393406

T2 - ACM/SIGDA Thirteenth ACM International Symposium on Field Programmable Gate Arrays - FPGA 2005

Y2 - 20 February 2005 through 22 February 2005

ER -

Time-multiplexed execution on the dynamically reconfigurable processor - A performance/cost evaluation

抄録

Conference

ASJC Scopus subject areas

他のファイルとリンク

フィンガープリント

引用スタイル