Estimation of energy consumption in machine learning

By Eva García-Martín, Crefeda Faviola Rodrigues, Graham D. Riley, Håkan Grahn. J. Parallel Distributed Comput., 2019

492 citations recorded by the source corpus.

Energy consumption has been widely studied in the computer architecture field for decades. While the adoption of energy as a metric in machine learning is emerging, the majority of research is stil ...

Subfields: Computer Networks and Communications, Electrical and Electronic Engineering, Hardware and Architecture

View Paper: https://www.semanticscholar.org/paper/d1faad340b63ac3c507701cb72ce5fef6c911c77

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

1. Luís Cruz
2. June Sallou
3. Luciano Floridi
4. Ben Hutchinson
5. Roberto Verdecchia

In Electrical and Electronic Engineering

1. Luís Cruz
2. Katherine A. Heller
3. June Sallou
4. Roberto Verdecchia

In Hardware and Architecture

1. Luciano Floridi
2. Negar Rostamzadeh
3. Josh Cowls

Citations in the corpus

• Unraveling the hidden environmental impacts of AI solutions for environment (2021, unknown): “Other authors [13] have also reviewed all the tools to estimate energy consumption from computer architecture.”
• Evaluation Gaps in Machine Learning Practice (2022, background): “[40, 60, 159] and Was the labour paid fairly?”
• The AI gambit: leveraging artificial intelligence to combat climate change—opportunities, challenges, and recommendations (2021, methodology): “…AI research projects may be indicative of — and exacerbate — a failure to engage with environmental questions, to disclose important research data, and to shift the focus away from ecologically short-sighted success metrics (García-Martín et al. 2019; Schwartz et al. 2019; Henderson et al. 2020).”
• The Computational Limits of Deep Learning (2020, background): “These projections reinforce the growing concern that deep learning’s current trajectory will have important negative effects [40, 79].”
• A systematic review of Green AI (2023, background): “Collecting and making sense of energy or climate data is far from trivial – many different factors contribute to the final estimation [48].”