Analysis of Increasing the Flexibility of LBs of Matrix-based FPGAs to Achieve Low Power Consumption

碩士 === 國立中山大學 === 電機工程研究所 === 85 === The large resistance and capacitance of the connection switches in Field Programmable Gate Arrays (FPGA) routing tracks consume a great portion of total power and cause large delays no matter what technology is used for the programmable connections. Hence,...

Full description

Bibliographic Details
Main Authors: Huang, Chia-Fu, 黃加富
Other Authors: Wang, Chua-Chin
Format: Others
Language:zh-TW
Published: 1997
Online Access:http://ndltd.ncl.edu.tw/handle/03352156139372311216
id ndltd-TW-085NSYS3442032
record_format oai_dc
spelling ndltd-TW-085NSYS34420322015-10-13T18:05:28Z http://ndltd.ncl.edu.tw/handle/03352156139372311216 Analysis of Increasing the Flexibility of LBs of Matrix-based FPGAs to Achieve Low Power Consumption 增加矩陣式場效可程式化邏輯陣列區塊接腳選擇度以達成低功率之分析 Huang, Chia-Fu 黃加富 碩士 國立中山大學 電機工程研究所 85 The large resistance and capacitance of the connection switches in Field Programmable Gate Arrays (FPGA) routing tracks consume a great portion of total power and cause large delays no matter what technology is used for the programmable connections. Hence, the reduction of the average number of the connection switches per net can lead to the power and delay reduction of the whole circuit. Consequently the performance of mapped circuits will be improved. We propose to double the pins of each input or output and arrange the two pins of each I/O symmetrically in the logic block (LB). An analytic general from solution of the reduction ratio of switch count is presented. Finally, thorough simulation results of certain benchmark circuits showing the impact on the power and delay are presented to verify the theoretical analysis. Wang, Chua-Chin 王朝欽 1997 學位論文 ; thesis 51 zh-TW
collection NDLTD
language zh-TW
format Others
sources NDLTD
description 碩士 === 國立中山大學 === 電機工程研究所 === 85 === The large resistance and capacitance of the connection switches in Field Programmable Gate Arrays (FPGA) routing tracks consume a great portion of total power and cause large delays no matter what technology is used for the programmable connections. Hence, the reduction of the average number of the connection switches per net can lead to the power and delay reduction of the whole circuit. Consequently the performance of mapped circuits will be improved. We propose to double the pins of each input or output and arrange the two pins of each I/O symmetrically in the logic block (LB). An analytic general from solution of the reduction ratio of switch count is presented. Finally, thorough simulation results of certain benchmark circuits showing the impact on the power and delay are presented to verify the theoretical analysis.
author2 Wang, Chua-Chin
author_facet Wang, Chua-Chin
Huang, Chia-Fu
黃加富
author Huang, Chia-Fu
黃加富
spellingShingle Huang, Chia-Fu
黃加富
Analysis of Increasing the Flexibility of LBs of Matrix-based FPGAs to Achieve Low Power Consumption
author_sort Huang, Chia-Fu
title Analysis of Increasing the Flexibility of LBs of Matrix-based FPGAs to Achieve Low Power Consumption
title_short Analysis of Increasing the Flexibility of LBs of Matrix-based FPGAs to Achieve Low Power Consumption
title_full Analysis of Increasing the Flexibility of LBs of Matrix-based FPGAs to Achieve Low Power Consumption
title_fullStr Analysis of Increasing the Flexibility of LBs of Matrix-based FPGAs to Achieve Low Power Consumption
title_full_unstemmed Analysis of Increasing the Flexibility of LBs of Matrix-based FPGAs to Achieve Low Power Consumption
title_sort analysis of increasing the flexibility of lbs of matrix-based fpgas to achieve low power consumption
publishDate 1997
url http://ndltd.ncl.edu.tw/handle/03352156139372311216
work_keys_str_mv AT huangchiafu analysisofincreasingtheflexibilityoflbsofmatrixbasedfpgastoachievelowpowerconsumption
AT huángjiāfù analysisofincreasingtheflexibilityoflbsofmatrixbasedfpgastoachievelowpowerconsumption
AT huangchiafu zēngjiājǔzhènshìchǎngxiàokěchéngshìhuàluójízhènlièqūkuàijiējiǎoxuǎnzédùyǐdáchéngdīgōnglǜzhīfēnxī
AT huángjiāfù zēngjiājǔzhènshìchǎngxiàokěchéngshìhuàluójízhènlièqūkuàijiējiǎoxuǎnzédùyǐdáchéngdīgōnglǜzhīfēnxī
_version_ 1718028052879900672