Module selection and data format conversion for cost-optimal DSP synthesis

Kazuhito Ito, Lori E. Lucke, Keshab K Parhi

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

In high level synthesis each node of a synchronous data-flow graph (DFG) is scheduled to a specific time and allocated to a processor. In this paper we present new integer linear programming (ILP) models which generate a blocked schedule for a DFG with implicit retiming, pipelining, and unfolding while performing module selection and data format conversion. A blocked schedule is a schedule which overlaps multiple iterations of the DFG to guarantee a minimum number of processors. Component modules are selected from a library of processors to minimize cost. Furthermore, we include data format converters between processors of different data formats. In addition, we minimize the unfolding factor of the blocked schedule.

Original languageEnglish (US)
Pages (from-to)322-329
Number of pages8
JournalIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
StatePublished - Dec 1 1994

Fingerprint

Data flow graphs
Costs
Linear programming

Cite this

Module selection and data format conversion for cost-optimal DSP synthesis. / Ito, Kazuhito; Lucke, Lori E.; Parhi, Keshab K.

In: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 01.12.1994, p. 322-329.

Research output: Contribution to journalArticle

@article{b18cf61db21e4d568fc7c05b47d299a3,
title = "Module selection and data format conversion for cost-optimal DSP synthesis",
abstract = "In high level synthesis each node of a synchronous data-flow graph (DFG) is scheduled to a specific time and allocated to a processor. In this paper we present new integer linear programming (ILP) models which generate a blocked schedule for a DFG with implicit retiming, pipelining, and unfolding while performing module selection and data format conversion. A blocked schedule is a schedule which overlaps multiple iterations of the DFG to guarantee a minimum number of processors. Component modules are selected from a library of processors to minimize cost. Furthermore, we include data format converters between processors of different data formats. In addition, we minimize the unfolding factor of the blocked schedule.",
author = "Kazuhito Ito and Lucke, {Lori E.} and Parhi, {Keshab K}",
year = "1994",
month = "12",
day = "1",
language = "English (US)",
pages = "322--329",
journal = "IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems",
issn = "0278-0070",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Module selection and data format conversion for cost-optimal DSP synthesis

AU - Ito, Kazuhito

AU - Lucke, Lori E.

AU - Parhi, Keshab K

PY - 1994/12/1

Y1 - 1994/12/1

N2 - In high level synthesis each node of a synchronous data-flow graph (DFG) is scheduled to a specific time and allocated to a processor. In this paper we present new integer linear programming (ILP) models which generate a blocked schedule for a DFG with implicit retiming, pipelining, and unfolding while performing module selection and data format conversion. A blocked schedule is a schedule which overlaps multiple iterations of the DFG to guarantee a minimum number of processors. Component modules are selected from a library of processors to minimize cost. Furthermore, we include data format converters between processors of different data formats. In addition, we minimize the unfolding factor of the blocked schedule.

AB - In high level synthesis each node of a synchronous data-flow graph (DFG) is scheduled to a specific time and allocated to a processor. In this paper we present new integer linear programming (ILP) models which generate a blocked schedule for a DFG with implicit retiming, pipelining, and unfolding while performing module selection and data format conversion. A blocked schedule is a schedule which overlaps multiple iterations of the DFG to guarantee a minimum number of processors. Component modules are selected from a library of processors to minimize cost. Furthermore, we include data format converters between processors of different data formats. In addition, we minimize the unfolding factor of the blocked schedule.

UR - http://www.scopus.com/inward/record.url?scp=0028714030&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028714030&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0028714030

SP - 322

EP - 329

JO - IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems

JF - IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems

SN - 0278-0070

ER -