Floorplan-Based Power/Ground Network Synthesis Using Lookup Tables
碩士 === 逢甲大學 === 資訊工程所 === 94 === In this paper, we present a novel design methodology to find an optimized floorplan and simultaneously establish its corresponding mesh-based power network so that the used power routing resource is minimized with considering IR voltage drop and electromigration (EM)...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2006
|
Online Access: | http://ndltd.ncl.edu.tw/handle/60144435425262420941 |
id |
ndltd-TW-094FCU05392030 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-094FCU053920302015-12-11T04:04:18Z http://ndltd.ncl.edu.tw/handle/60144435425262420941 Floorplan-Based Power/Ground Network Synthesis Using Lookup Tables 使用查表法的平面規劃電源供應網路合成 Tai-wei Kung 龔泰維 碩士 逢甲大學 資訊工程所 94 In this paper, we present a novel design methodology to find an optimized floorplan and simultaneously establish its corresponding mesh-based power network so that the used power routing resource is minimized with considering IR voltage drop and electromigration (EM). This proposed methodology breaks the traditional physical design procedure by building power network prior to deciding floorplanning. A novel and efficient regular power network table is created to provide gradient information for fast signal-integrity (SI) estimation, which results in efficient solution exploration. Based on the concept of genetic algorithm, a fixed-outline driven floorplanning algorithm is applied to explore floorplans to gradually satisfy the SI issues. The proposed flow not only significantly speeds up the optimization process, without repeatedly calculating complex matrix inverse, but also shortens the turn-around time from floorplanning to power analysis of a design. De-sheng Chen 陳德生 2006 學位論文 ; thesis 53 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 逢甲大學 === 資訊工程所 === 94 === In this paper, we present a novel design methodology to find an optimized floorplan and simultaneously establish its corresponding mesh-based power network so that the used power routing resource is minimized with considering IR voltage drop and electromigration (EM). This proposed methodology breaks the traditional physical design procedure by building power network prior to deciding floorplanning. A novel and efficient regular power network table is created to provide gradient information for fast signal-integrity (SI) estimation, which results in efficient solution exploration. Based on the concept of genetic algorithm, a fixed-outline driven floorplanning algorithm is applied to explore floorplans to gradually satisfy the SI issues. The proposed flow not only significantly speeds up the optimization process, without repeatedly calculating complex matrix inverse, but also shortens the turn-around time from floorplanning to power analysis of a design.
|
author2 |
De-sheng Chen |
author_facet |
De-sheng Chen Tai-wei Kung 龔泰維 |
author |
Tai-wei Kung 龔泰維 |
spellingShingle |
Tai-wei Kung 龔泰維 Floorplan-Based Power/Ground Network Synthesis Using Lookup Tables |
author_sort |
Tai-wei Kung |
title |
Floorplan-Based Power/Ground Network Synthesis Using Lookup Tables |
title_short |
Floorplan-Based Power/Ground Network Synthesis Using Lookup Tables |
title_full |
Floorplan-Based Power/Ground Network Synthesis Using Lookup Tables |
title_fullStr |
Floorplan-Based Power/Ground Network Synthesis Using Lookup Tables |
title_full_unstemmed |
Floorplan-Based Power/Ground Network Synthesis Using Lookup Tables |
title_sort |
floorplan-based power/ground network synthesis using lookup tables |
publishDate |
2006 |
url |
http://ndltd.ncl.edu.tw/handle/60144435425262420941 |
work_keys_str_mv |
AT taiweikung floorplanbasedpowergroundnetworksynthesisusinglookuptables AT gōngtàiwéi floorplanbasedpowergroundnetworksynthesisusinglookuptables AT taiweikung shǐyòngchábiǎofǎdepíngmiànguīhuàdiànyuángōngyīngwǎnglùhéchéng AT gōngtàiwéi shǐyòngchábiǎofǎdepíngmiànguīhuàdiànyuángōngyīngwǎnglùhéchéng |
_version_ |
1718147572773683200 |