Mitigating Curtailment and Carbon Emissions through Load Migration between Data Centers

Status:: πŸŸ₯
Links:: Carbon-Aware Computing

Metadata

Authors:: Zheng, Jiajia; Chien, Andrew A.; Suh, Sangwon
Title:: Mitigating Curtailment and Carbon Emissions through Load Migration between Data Centers
Publication Title:: "Joule"
Date:: 2020
URL:: https://www.cell.com/joule/abstract/S2542-4351(20)30347-0
DOI:: 10.1016/j.joule.2020.08.001

Bibliography

Zheng, J., Chien, A. A., & Suh, S. (2020). Mitigating Curtailment and Carbon Emissions through Load Migration between Data Centers. Joule, 4(10), 2208–2222. https://doi.org/10.1016/j.joule.2020.08.001

Zotero

Type:: #zotero/journalArticle
Zotero::

Keywords:: [/unread, curtailment, data center, greenhouse gas emissions mitigation, load migration, variable renewable energy]

Relations

Notes & Annotations

🟨 Note (last modified: 2023-08-23#09:22:15) (

Paper notes by David Mytton

https://davidmytton.blog/paper-notes-mitigating-curtailment-and-carbon-emissions-through-load-migration-between-data-centers/

This is a good paper that makes valid points about the possibilities of migrating flexible IT workloads, however it makes classic assumptions I see in most papers that discuss this topic. The big assumptions are that the data centers are fungible entities where the operator has complete control, the goal is to optimize for efficiency, and it is trivial to move huge volumes of data cheaply.

We are starting to see this with the big tech companies who own many data centers. Google is an example used in the paper, but it’s telling that it is the only example.Β Google is only doing it for a small number of workloads. That shows how difficult it is.

This paper is good inspiration for future research direction. I think we’ve seen enough evidence of the benefits of such capabilities that time and effort should be directed towards building the technologies needed to achieve these migrations. And if the cloud providersΒ could reduce their data transfer fees, that would be nice as well.