University of Twente Student Theses

Login

T-SCARS : A Time-Sensitive, Spatial Carbon-Aware Request Scheduler for Data Center Networks

Smit, Anna (2025) T-SCARS : A Time-Sensitive, Spatial Carbon-Aware Request Scheduler for Data Center Networks.

[img] PDF
610kB
Abstract:Data centers consumed in 2022, 2% of the world’s energy supply and are estimated to grow in terms of energy consumption and therefore, carbon emissions. Technology companies have recognized this and set aggressive targets for reducing their carbon footprint. This called for new research methods that directly target the reduction of carbon emissions, one of which is spatial workload-scheduling. This approach shifts the carbon-intense computations across space, to where lower-carbon intensive energy is available. This study delves into the foundations of a spatial-aware algorithm by developing T-SCARS, a time-sensitive, spatial carbon-aware request scheduler. T-SCARS could be promising due to its priority to schedule requests to carbon-efficient data centers whilst keeping to a time constraint of transferring and processing the request. We evaluated the impact of T-SCARS on carbon emissions through various simulations by modelling a data center network and comparing the results against carbon-oblivious baselines. These experiments involve scaling the model of the data center network, increasing the number of requests, adjusting the timeliness of the requests, and analyzing T-SCARS’s consistency. The results highlight the potential of T-SCARS, which consistently outperforms its carbon-oblivious counterparts by achieving lower carbon emissions, often resulting in only a slight increase in time-constraint violations.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Computer Science BSc (56964)
Link to this item:https://purl.utwente.nl/essays/105115
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page