A low-power digital filter IC via soft DSP

R. Hegde, N. R. Shanbhag

Research output: Contribution to journalConference article

Abstract

In this paper we present an integrated circuit implementation of a soft DSP based low-power digital filter in 0.35μm, 3.3V CMOS process. Soft DSP is a low-power technique that employs voltage overscaling (VOS) and algorithmic noise-tolerance (ANT) to push the limits of energy-efficiency beyond that achievable by voltage scaling along. VOS refers to scaling the supply voltage beyond the limit imposed by the throughput constraints. ANT is an algorithmic level error-control technique that is employed to restore the algorithmic performance degradation in terms of output signal-to-noise ratio (SNR) caused by VOS. Measured results indicate 40% - 67% reduction in energy dissipation over optimally voltage-scaled systems with less than 1db loss in S N R for a wide range of filter bandwidths (0.05fs - 0.25fs, where fs is the sampling frequency).

Original languageEnglish (US)
Pages (from-to)309-312
Number of pages4
JournalProceedings of the Custom Integrated Circuits Conference
StatePublished - Jan 1 2001
EventIEEE 2001 Custom Integrated Circuits Conference - San Diego, CA, United States
Duration: May 6 2001May 9 2001

Fingerprint

Digital filters
Electric potential
Integrated circuits
Energy efficiency
Energy dissipation
Signal to noise ratio
Throughput
Sampling
Bandwidth
Degradation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

A low-power digital filter IC via soft DSP. / Hegde, R.; Shanbhag, N. R.

In: Proceedings of the Custom Integrated Circuits Conference, 01.01.2001, p. 309-312.

Research output: Contribution to journalConference article

@article{9318b0e00e614cea9c7f6576166a698a,
title = "A low-power digital filter IC via soft DSP",
abstract = "In this paper we present an integrated circuit implementation of a soft DSP based low-power digital filter in 0.35μm, 3.3V CMOS process. Soft DSP is a low-power technique that employs voltage overscaling (VOS) and algorithmic noise-tolerance (ANT) to push the limits of energy-efficiency beyond that achievable by voltage scaling along. VOS refers to scaling the supply voltage beyond the limit imposed by the throughput constraints. ANT is an algorithmic level error-control technique that is employed to restore the algorithmic performance degradation in terms of output signal-to-noise ratio (SNR) caused by VOS. Measured results indicate 40{\%} - 67{\%} reduction in energy dissipation over optimally voltage-scaled systems with less than 1db loss in S N R for a wide range of filter bandwidths (0.05fs - 0.25fs, where fs is the sampling frequency).",
author = "R. Hegde and Shanbhag, {N. R.}",
year = "2001",
month = "1",
day = "1",
language = "English (US)",
pages = "309--312",
journal = "Proceedings of the Custom Integrated Circuits Conference",
issn = "0886-5930",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - A low-power digital filter IC via soft DSP

AU - Hegde, R.

AU - Shanbhag, N. R.

PY - 2001/1/1

Y1 - 2001/1/1

N2 - In this paper we present an integrated circuit implementation of a soft DSP based low-power digital filter in 0.35μm, 3.3V CMOS process. Soft DSP is a low-power technique that employs voltage overscaling (VOS) and algorithmic noise-tolerance (ANT) to push the limits of energy-efficiency beyond that achievable by voltage scaling along. VOS refers to scaling the supply voltage beyond the limit imposed by the throughput constraints. ANT is an algorithmic level error-control technique that is employed to restore the algorithmic performance degradation in terms of output signal-to-noise ratio (SNR) caused by VOS. Measured results indicate 40% - 67% reduction in energy dissipation over optimally voltage-scaled systems with less than 1db loss in S N R for a wide range of filter bandwidths (0.05fs - 0.25fs, where fs is the sampling frequency).

AB - In this paper we present an integrated circuit implementation of a soft DSP based low-power digital filter in 0.35μm, 3.3V CMOS process. Soft DSP is a low-power technique that employs voltage overscaling (VOS) and algorithmic noise-tolerance (ANT) to push the limits of energy-efficiency beyond that achievable by voltage scaling along. VOS refers to scaling the supply voltage beyond the limit imposed by the throughput constraints. ANT is an algorithmic level error-control technique that is employed to restore the algorithmic performance degradation in terms of output signal-to-noise ratio (SNR) caused by VOS. Measured results indicate 40% - 67% reduction in energy dissipation over optimally voltage-scaled systems with less than 1db loss in S N R for a wide range of filter bandwidths (0.05fs - 0.25fs, where fs is the sampling frequency).

UR - http://www.scopus.com/inward/record.url?scp=0034829115&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034829115&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:0034829115

SP - 309

EP - 312

JO - Proceedings of the Custom Integrated Circuits Conference

JF - Proceedings of the Custom Integrated Circuits Conference

SN - 0886-5930

ER -