Adaptive Approximate Computing in Arithmetic Datapaths

被引：14

作者：

Mazahir, Sana ^{[1
]}

Hasan, Osman ^{[2
]}

Shafique, Muhammad ^{[3
]}

机构：

[1] Georgia Inst Technol, Atlanta, GA 30332 USA

[2] Natl Univ Sci & Technol, Islamabad, Pakistan

[3] Vienna Univ Technol, Inst Comp Engn, Vienna, Austria

来源：

IEEE DESIGN & TEST | 2018年 / 35卷 / 04期

关键词：

accuracy configurability; adaptive computing; approximate adders; Approximate computing; approximate multipliers; error detection and correction (EDC); self-aware architectures;

D O I：

10.1109/MDAT.2017.2772874

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

Approximate arithmetic units are used in error tolerant applications, like image processing, to minimize energy, and resource usage in general, by keeping accuracy to the minimum level required. However, in a datapath errors may accumulate over successive approximate operations. This article proposes a scheme to bound the size of the errors with adaptive error detection and correction as an integral part of the datapath.

引用

页码：65 / 74

页数：10

共 50 条

[21] Dynamically Specialized Datapaths for Energy Efficient Computing
Govindaraju, Venkatraman
Ho, Chen-Han
Sankaralingam, Karthikeyan
2011 IEEE 17TH INTERNATIONAL SYMPOSIUM ON HIGH-PERFORMANCE COMPUTER ARCHITECTURE (HPCA), 2011, : 503 - 514
[22] High performance approximate computing by adaptive relaxed synchronization
Islam, Bashima
Shezan, Faysal Hossain
Shahriyar, Rifat
Proceedings - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016, 2017, : 1204 - 1210
[23] High Performance Approximate Computing by Adaptive Relaxed Synchronization
Islam, Bashima
Shezan, Faysal Hossain
Shahriyar, Rifat
PROCEEDINGS OF 2016 IEEE 18TH INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING AND COMMUNICATIONS; IEEE 14TH INTERNATIONAL CONFERENCE ON SMART CITY; IEEE 2ND INTERNATIONAL CONFERENCE ON DATA SCIENCE AND SYSTEMS (HPCC/SMARTCITY/DSS), 2016, : 1204 - 1210
[24] Stateful Adaptive Streams with Approximate Computing and Elastic Scaling
Francisco, Joao
Coimbra, Miguel E.
Neto, Pedro F.
Freitag, Felix
Veiga, Luis
38TH ANNUAL ACM SYMPOSIUM ON APPLIED COMPUTING, SAC 2023, 2023, : 174 - 183
[25] ACOCO: Adaptive Coding for Approximate Computing on Faulty Memories
Huang, Chu-Hsiang
Li, Yao
Dolecek, Lara
IEEE TRANSACTIONS ON COMMUNICATIONS, 2015, 63 (12) : 4615 - 4628
[26] Prediction of the Impact of Approximate Computing on Spiking Neural Networks via Interval Arithmetic
Saeedi, Sepide
Carpegna, Alessio
Savino, Alessandro
Di Carlo, Stefano
2022 23RD IEEE LATIN-AMERICAN TEST SYMPOSIUM (LATS 2022), 2022,
[27] Optimization of arithmetic datapaths with finite word-length operands
Gopalakrishnan, Sivaram
Kalla, Priyank
Enescu, Florian
PROCEEDINGS OF THE ASP-DAC 2007, 2007, : 511 - +
[28] Retiming Arithmetic Datapaths using Timed Taylor Expansion Diagrams
Gomez-Prado, Daniel
Kim, Dusung
Ciesielski, Maciej
Boutillon, Emmanuel
2010 IEEE INTERNATIONAL HIGH LEVEL DESIGN VALIDATION AND TEST WORKSHOP (HLDVT), 2010, : 33 - 39
[29] Computing multiple roots of polynomials in stochastic arithmetic with Newton method and approximate GCD
Graillat, Stef
Jezequel, Fabienne
Martins, Enzo Queiros
Spyropoulos, Maxime
2021 23RD INTERNATIONAL SYMPOSIUM ON SYMBOLIC AND NUMERIC ALGORITHMS FOR SCIENTIFIC COMPUTING (SYNASC 2021), 2021, : 45 - 51
[30] A Fast and Fuzzy Functional Simulator of Inexact Arithmetic Operators for Approximate Computing Systems
Bonnot, Justine
Desnos, Karol
Pelcat, Maxime
Menard, Daniel
PROCEEDINGS OF THE 2018 GREAT LAKES SYMPOSIUM ON VLSI (GLSVLSI'18), 2018, : 195 - 200

← 1 2 3 4 5 →