Reliable low-power digital signal processing via reduced precision redundancy

Byonghyo Shim; Srinivasa R. Sridhara; Naresh R. Shanbhag

doi:10.1109/TVLSI.2004.826201

Reliable low-power digital signal processing via reduced precision redundancy

Byonghyo Shim, Srinivasa R. Sridhara, Naresh R. Shanbhag

Department of Computer Science and Engineering

Research output: Contribution to journal › Article › peer-review

153 Citations (Scopus)

Abstract

In this paper, we present a novel algorithmic noise-tolerance (ANT) technique referred to as reduced precision redundancy (RPR). RPR requires a reduced precision replica whose output can be employed as the corrected output in case the original system computes erroneously. When combined with voltage overscaling (VOS), the resulting soft digital signal processing system achieves up to 60% and 44% energy savings with no loss in the signal-to-noise ratio (SNR) for receive filtering in a QPSK system and the butterfly of fast Fourier transform (FFT) in a WLAN OFDM system, respectively. These energy savings are with respect to optimally scaled (i.e., the supply voltage equals the critical voltage V_{dd-cr it}) present day systems. Further, we show that the RPR technique is able to maintain the output SNR for error rates of up to 0.09/sample and 0.06/sample in an finite impulse response filter and a FFT block, respectively.

Original language	English
Pages (from-to)	497-510
Number of pages	14
Journal	IEEE Transactions on Very Large Scale Integration (VLSI) Systems
Volume	12
Issue number	5
DOIs	https://doi.org/10.1109/TVLSI.2004.826201
Publication status	Published - 2004 May

Keywords

Digital signal processing
Low-power
Noise-tolerance
Reliability
Supply voltage

ASJC Scopus subject areas

Software
Hardware and Architecture
Electrical and Electronic Engineering

Access to Document

10.1109/TVLSI.2004.826201

Cite this

@article{23ff3778e18d4cc5877eadb680e6b919,

title = "Reliable low-power digital signal processing via reduced precision redundancy",

abstract = "In this paper, we present a novel algorithmic noise-tolerance (ANT) technique referred to as reduced precision redundancy (RPR). RPR requires a reduced precision replica whose output can be employed as the corrected output in case the original system computes erroneously. When combined with voltage overscaling (VOS), the resulting soft digital signal processing system achieves up to 60% and 44% energy savings with no loss in the signal-to-noise ratio (SNR) for receive filtering in a QPSK system and the butterfly of fast Fourier transform (FFT) in a WLAN OFDM system, respectively. These energy savings are with respect to optimally scaled (i.e., the supply voltage equals the critical voltage Vdd-cr it) present day systems. Further, we show that the RPR technique is able to maintain the output SNR for error rates of up to 0.09/sample and 0.06/sample in an finite impulse response filter and a FFT block, respectively.",

keywords = "Digital signal processing, Low-power, Noise-tolerance, Reliability, Supply voltage",

author = "Byonghyo Shim and Sridhara, {Srinivasa R.} and Shanbhag, {Naresh R.}",

note = "Funding Information: Manuscript received May 2, 2003; revised September 29, 2003. This research was supported in part by the Microelectronics Advanced Research Corporation (MARCO) sponsored Gigascale Silicon Research Center and National Science Foundation under Grant CCR 99-79381 and Grant CCR 00-85929.",

year = "2004",

month = may,

doi = "10.1109/TVLSI.2004.826201",

language = "English",

volume = "12",

pages = "497--510",

journal = "IEEE Transactions on Very Large Scale Integration (VLSI) Systems",

issn = "1063-8210",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "5",

}

TY - JOUR

T1 - Reliable low-power digital signal processing via reduced precision redundancy

AU - Shim, Byonghyo

AU - Sridhara, Srinivasa R.

AU - Shanbhag, Naresh R.

N1 - Funding Information: Manuscript received May 2, 2003; revised September 29, 2003. This research was supported in part by the Microelectronics Advanced Research Corporation (MARCO) sponsored Gigascale Silicon Research Center and National Science Foundation under Grant CCR 99-79381 and Grant CCR 00-85929.

PY - 2004/5

Y1 - 2004/5

N2 - In this paper, we present a novel algorithmic noise-tolerance (ANT) technique referred to as reduced precision redundancy (RPR). RPR requires a reduced precision replica whose output can be employed as the corrected output in case the original system computes erroneously. When combined with voltage overscaling (VOS), the resulting soft digital signal processing system achieves up to 60% and 44% energy savings with no loss in the signal-to-noise ratio (SNR) for receive filtering in a QPSK system and the butterfly of fast Fourier transform (FFT) in a WLAN OFDM system, respectively. These energy savings are with respect to optimally scaled (i.e., the supply voltage equals the critical voltage Vdd-cr it) present day systems. Further, we show that the RPR technique is able to maintain the output SNR for error rates of up to 0.09/sample and 0.06/sample in an finite impulse response filter and a FFT block, respectively.

AB - In this paper, we present a novel algorithmic noise-tolerance (ANT) technique referred to as reduced precision redundancy (RPR). RPR requires a reduced precision replica whose output can be employed as the corrected output in case the original system computes erroneously. When combined with voltage overscaling (VOS), the resulting soft digital signal processing system achieves up to 60% and 44% energy savings with no loss in the signal-to-noise ratio (SNR) for receive filtering in a QPSK system and the butterfly of fast Fourier transform (FFT) in a WLAN OFDM system, respectively. These energy savings are with respect to optimally scaled (i.e., the supply voltage equals the critical voltage Vdd-cr it) present day systems. Further, we show that the RPR technique is able to maintain the output SNR for error rates of up to 0.09/sample and 0.06/sample in an finite impulse response filter and a FFT block, respectively.

KW - Digital signal processing

KW - Low-power

KW - Noise-tolerance

KW - Reliability

KW - Supply voltage

UR - http://www.scopus.com/inward/record.url?scp=2542436055&partnerID=8YFLogxK

U2 - 10.1109/TVLSI.2004.826201

DO - 10.1109/TVLSI.2004.826201

M3 - Article

AN - SCOPUS:2542436055

SN - 1063-8210

VL - 12

SP - 497

EP - 510

JO - IEEE Transactions on Very Large Scale Integration (VLSI) Systems

JF - IEEE Transactions on Very Large Scale Integration (VLSI) Systems

IS - 5

ER -

Reliable low-power digital signal processing via reduced precision redundancy

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this