Search: onr:"swepub:oai:DiVA.org:mdh-47029" >
ΔnN :
ΔnN : Power-Efficient Neural Network Acceleration Using Differential Weights
-
- Mahdiani, Hoda (author)
- Univ Tehran, Dept Elect & Comp Engn, Comp Engn, Tehran, Iran.
-
- Khadem, Alireza (author)
- Univ Tehran, Tehran, Iran.
-
- Ghanbari, Azam (author)
- Univ Tehran, Dept Elect & Comp Engn, Comp Engn, Tehran, Iran.
-
show more...
-
- Modarressi, Mehdi (author)
- Univ Tehran, Coll Engn, Dept Elect & Comp Engn, Tehran, Iran.
-
- Fattahi-Bayat, Farima (author)
- Univ Tehran, Tehran, Iran.
-
- Daneshtalab, Masoud (author)
- Mälardalens högskola,Inbyggda system
-
show less...
-
Univ Tehran, Dept Elect & Comp Engn, Comp Engn, Tehran, Iran Univ Tehran, Tehran, Iran. (creator_code:org_t)
- IEEE COMPUTER SOC, 2020
- 2020
- English.
-
In: IEEE Micro. - : IEEE COMPUTER SOC. - 0272-1732 .- 1937-4143. ; 40:1, s. 67-74
- Related links:
-
https://urn.kb.se/re...
-
show more...
-
https://doi.org/10.1...
-
show less...
Abstract
Subject headings
Close
- The enormous and ever-increasing complexity of state-of-the-art neural networks has impeded the deployment of deep learning on resource-limited embedded and mobile devices. To reduce the complexity of neural networks, this article presents Delta NN, a power-efficient architecture that leverages a combination of the approximate value locality of neuron weights and algorithmic structure of neural networks. Delta NN keeps each weight as its difference (Delta) to the nearest smaller weight: each weight reuses the calculations of the smaller weight, followed by a calculation on the Delta value to make up the difference. We also round up/down the Delta to the closest power of two numbers to further reduce complexity. The experimental results show that Delta NN boosts the average performance by 14%-37% and reduces the average power consumption by 17%-49% over some state-of-the-art neural network designs.
Subject headings
- TEKNIK OCH TEKNOLOGIER -- Elektroteknik och elektronik -- Kommunikationssystem (hsv//swe)
- ENGINEERING AND TECHNOLOGY -- Electrical Engineering, Electronic Engineering, Information Engineering -- Communication Systems (hsv//eng)
Publication and Content Type
- ref (subject category)
- art (subject category)
Find in a library
-
IEEE Micro
(Search for host publication in LIBRIS)
To the university's database