print

DeepAxe : a framework for exploration of approximation and reliability trade-offs in DNN accelerators

author
Taheri, Mahdi
Riazati, Mohamad
Ahmadilivani, Mohammad Hasan
Jenihhin, Maksim
Daneshtalab, Masoud
Raik, Jaan
Sjödin, Mikael
Lisper, Björn
statement of authorship
Mahdi Taheri, Mohammad Riazati, Mohammad Hasan Ahmadilivani, Maksim Jenihhin, Masoud Daneshtalab, Jaan Raik, Mikael Sjodin, Bjorn Lisper
source
arXiv.org
publisher
IEEE
journal volume number month
arXiv:2303.08226
year of publication
2023
pages
8 p. : ill
url
https://doi.org/10.48550/arXiv.2303.0822
subject term
testimine
rikked
kompuutersimulatsioon
tehisnärvivõrgud
keyword
deep neural networks
approximate computing
fault simulation
reliability
resiliency assessment
notes
This paper is accepted at the 24th International Symposium on Quality Electronic Design (ISQED) 2023
scientific publication
teaduspublikatsioon
classifier
3.1
TTÜ department
arvutisüsteemide instituut
language
inglise
Uurimisrühm
Centre for trustworthy and efficient computing hardware (TECH)
Taheri, M., Riazati, M., Ahmadilivani, M.H., Jenihhin, M., Daneshtalab, M., Raik, J., Sjödin, M., Lisper, B. DeepAxe : a framework for exploration of approximation and reliability trade-offs in DNN accelerators // arXiv.org (2023) arXiv:2303.08226, 8 p. : ill. https://doi.org/10.48550/arXiv.2303.0822