Taheri, M., Cherezova, N., Nazari, S., Azarpeyvand, A., Ghasempouri, T., Daneshtalab, M., Raik, J., Jenihhin, M.

full record print

AdAM: Adaptive Approximate Multiplier for Fault Tolerance in DNN Accelerators

author

Taheri, Mahdi

Cherezova, Natalia

Nazari, Samira

Azarpeyvand, Ali

Ghasempouri, Tara

Daneshtalab, Masoud

Raik, Jaan

Jenihhin, Maksim

statement of authorship

Mahdi Taheri, Natalia Cherezova, Samira Nazari, Ali Azarpeyvand, Tara Ghasempouri, Masoud Daneshtalab, Jaan Raik, and Maksim Jenihhin

source

IEEE transactions on device and materials reliability

publisher

IEEE

journal volume number month

vol. 25, no. 1

year of publication

2024

pages

p. 66-75 : ill

url

https://doi.org//10.1109/TDMR.2024.3523386

ISSN

1530-4388

notes

Bibliogr.: 43 ref

scientific publication

teaduspublikatsioon

language

inglise

Märksõnad ja võtmesõnad

subject term

tehisnärvivõrgud

kiirendid

elektriahelad

tõrketaluvus

keyword

Deep neural networks

approximate computing

circuit design

reliability

DNN accelerator

Teaduspublikatsiooni meetrika

classifier

1.1

TalTechi üksused

TalTech department

arvutisüsteemide instituut

Related publications

Methods for reliability assessment and enhancement of deep neural networks hardware accelerators = Süvanärvivõrkude riistvara kiirendite töökindluse hindamine ja täiustamine

Taheri, M., Cherezova, N., Nazari, S., Azarpeyvand, A., Ghasempouri, T., Daneshtalab, M., Raik, J., Jenihhin, M. AdAM: Adaptive Approximate Multiplier for Fault Tolerance in DNN Accelerators // IEEE transactions on device and materials reliability (2024) vol. 25, no. 1, p. 66-75 : ill. https://doi.org//10.1109/TDMR.2024.3523386