Saved in:
Bibliographic Details
Main Authors: Asperti, Andrea, Evangelista, Davide, Marzolla, Moreno
Format: Preprint
Published: 2021
Subjects:
Online Access:https://arxiv.org/abs/2107.11949
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912688328671232
author Asperti, Andrea
Evangelista, Davide
Marzolla, Moreno
author_facet Asperti, Andrea
Evangelista, Davide
Marzolla, Moreno
contents The term GreenAI refers to a novel approach to Deep Learning, that is more aware of the ecological impact and the computational efficiency of its methods. The promoters of GreenAI suggested the use of Floating Point Operations (FLOPs) as a measure of the computational cost of Neural Networks; however, that measure does not correlate well with the energy consumption of hardware equipped with massively parallel processing units like GPUs or TPUs. In this article, we propose a simple refinement of the formula used to compute floating point operations for convolutional layers, called α-FLOPs, explaining and correcting the traditional discrepancy with respect to different layers, and closer to reality. The notion of α-FLOPs relies on the crucial insight that, in case of inputs with multiple dimensions, there is no reason to believe that the speedup offered by parallelism will be uniform along all different axes.
format Preprint
id arxiv_https___arxiv_org_abs_2107_11949
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle Dissecting FLOPs along input dimensions for GreenAI cost estimations
Asperti, Andrea
Evangelista, Davide
Marzolla, Moreno
Machine Learning
68T07
I.2
The term GreenAI refers to a novel approach to Deep Learning, that is more aware of the ecological impact and the computational efficiency of its methods. The promoters of GreenAI suggested the use of Floating Point Operations (FLOPs) as a measure of the computational cost of Neural Networks; however, that measure does not correlate well with the energy consumption of hardware equipped with massively parallel processing units like GPUs or TPUs. In this article, we propose a simple refinement of the formula used to compute floating point operations for convolutional layers, called α-FLOPs, explaining and correcting the traditional discrepancy with respect to different layers, and closer to reality. The notion of α-FLOPs relies on the crucial insight that, in case of inputs with multiple dimensions, there is no reason to believe that the speedup offered by parallelism will be uniform along all different axes.
title Dissecting FLOPs along input dimensions for GreenAI cost estimations
topic Machine Learning
68T07
I.2
url https://arxiv.org/abs/2107.11949