Toward Dynamic Precision Scaling

Serif Yesil; Ismail Akturk; Ulya R. Karpuzcu

doi:10.1109/MM.2018.043191123

Toward Dynamic Precision Scaling

Serif Yesil, Ismail Akturk, Ulya R. Karpuzcu

Electrical and Computer Engineering

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

12 Scopus citations

Abstract

This article makes the case for dynamic precision scaling to improve power efficiency by tailoring arithmetic precision adaptively to temporal changes in algorithmic noise tolerance throughout the execution.

Original language	English (US)
Article number	8430621
Pages (from-to)	30-39
Number of pages	10
Journal	IEEE Micro
Volume	38
Issue number	4
DOIs	https://doi.org/10.1109/MM.2018.043191123
State	Published - Jul 1 2018

Bibliographical note

Funding Information:
This work was supported in part by NSF grant no. CCF-1438286.

Publisher Copyright:
© 1981-2012 IEEE.

Keywords

dynamic precision scaling
floating point arithmetic
hardware
precision reduction

Publisher link

10.1109/MM.2018.043191123

Find It

Find It at U of M Libraries

Cite this

@article{a64bfeb5a25d48df977242027288ddb0,

title = "Toward Dynamic Precision Scaling",

abstract = "This article makes the case for dynamic precision scaling to improve power efficiency by tailoring arithmetic precision adaptively to temporal changes in algorithmic noise tolerance throughout the execution.",

keywords = "dynamic precision scaling, floating point arithmetic, hardware, precision reduction",

author = "Serif Yesil and Ismail Akturk and Karpuzcu, {Ulya R.}",

note = "Funding Information: This work was supported in part by NSF grant no. CCF-1438286. Publisher Copyright: {\textcopyright} 1981-2012 IEEE.",

year = "2018",

month = jul,

day = "1",

doi = "10.1109/MM.2018.043191123",

language = "English (US)",

volume = "38",

pages = "30--39",

journal = "IEEE Micro",

issn = "0272-1732",

publisher = "IEEE Computer Society",

number = "4",

}

TY - JOUR

T1 - Toward Dynamic Precision Scaling

AU - Yesil, Serif

AU - Akturk, Ismail

AU - Karpuzcu, Ulya R.

PY - 2018/7/1

Y1 - 2018/7/1

N2 - This article makes the case for dynamic precision scaling to improve power efficiency by tailoring arithmetic precision adaptively to temporal changes in algorithmic noise tolerance throughout the execution.

AB - This article makes the case for dynamic precision scaling to improve power efficiency by tailoring arithmetic precision adaptively to temporal changes in algorithmic noise tolerance throughout the execution.

KW - dynamic precision scaling

KW - floating point arithmetic

KW - hardware

KW - precision reduction

UR - http://www.scopus.com/inward/record.url?scp=85051562253&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051562253&partnerID=8YFLogxK

U2 - 10.1109/MM.2018.043191123

DO - 10.1109/MM.2018.043191123

M3 - Article

AN - SCOPUS:85051562253

SN - 0272-1732

VL - 38

SP - 30

EP - 39

JO - IEEE Micro

JF - IEEE Micro

IS - 4

M1 - 8430621

ER -

Toward Dynamic Precision Scaling

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this