Dr. Brian Setz

• e-mail: briansetz [at] gmail.com
• personal website: Brian Setz

Research

• Service-Oriented Architecture
• Wireless Sensor Networks
• Data Center Efficiency

Recent publications

1. CO2 Emission Aware Scheduling for Deep Neural Network Training Workloads (K. Haghshenas, B. Setz, and M. Aiello), In 2022 IEEE International Conference on Big Data (Big Data), IEEE, 2022.

   Abstract

   Machine Learning (ML) training is a growing workload in high-performance computing clusters and data centers; furthermore, it is computationally intensive and requires substantial amounts of energy with associated emissions. To the best of our knowledge, previous works in the area of load management have never focused on decreasing the carbon emission of ML training workloads. In this paper, we explore the potential emission reduction achievable by leveraging the iterative nature of the training process as well as the variability of CO 2 signal intensity as coming from the power grid. To this end, we introduce two emission-aware mechanisms to shift the training jobs in time and migrate them between geographical locations. We present experimental results on power and carbon emission of the training process together with delay overheads associated with emission reduction mechanisms, for various, representative, deep neural network models. The results show that following emission signals, one can effectively reduce emissions by an amount that varies from 13% to 57% of the baseline cases. Moreover, the experimental results show that the total delay overhead for applying emission-aware mechanisms multiple times is negligible compared to the jobs’ completion time.

   ---

   BibTeX

   
   @inproceedings{haghshenas2022:bd, title = "CO2 Emission Aware Scheduling for Deep Neural Network Training Workloads", abstract = "Machine Learning (ML) training is a growing workload in high-performance computing clusters and data centers; furthermore, it is computationally intensive and requires substantial amounts of energy with associated emissions. To the best of our knowledge, previous works in the area of load management have never focused on decreasing the carbon emission of ML training workloads. In this paper, we explore the potential emission reduction achievable by leveraging the iterative nature of the training process as well as the variability of CO 2 signal intensity as coming from the power grid. To this end, we introduce two emission-aware mechanisms to shift the training jobs in time and migrate them between geographical locations. We present experimental results on power and carbon emission of the training process together with delay overheads associated with emission reduction mechanisms, for various, representative, deep neural network models. The results show that following emission signals, one can effectively reduce emissions by an amount that varies from 13% to 57% of the baseline cases. Moreover, the experimental results show that the total delay overhead for applying emission-aware mechanisms multiple times is negligible compared to the jobs’ completion time.", author = "K. Haghshenas and B. Setz, and M. Aiello", year = "2022", % month = dec, language = "English", booktitle = "2022 IEEE International Conference on Big Data (Big Data)", publisher = "IEEE", doi = "10.1109/BigData55660.2022.10020544", url = "https://ieeexplore.ieee.org/abstract/document/10020544", }

   ---

   urldoi
2. Mining Sequential Patterns for Appliance Usage Prediction (M. Kalksma, B. Setz, A. R. Pratama, I. Georgievski and M. Aiello), In International Conference on Smart Cities and Green ICT Systems, 2018.

   BibTeX

   
   @inproceedings{kalksma2018:mining, title = {Mining Sequential Patterns for Appliance Usage Prediction}, author = {M. Kalksma and B. Setz and A. R. Pratama and I. Georgievski and M. Aiello}, booktitle = {International Conference on Smart Cities and Green ICT Systems}, series = {SMARTGREENS'18}, year = {2018}, note = {To appear}, }

   ---

3. Planning meets activity recognition: Service coordination for intelligent buildings (I. Georgievski, T. A. Nguyen, F. Nizamic, B. Setz, A. Lazovik and M. Aiello), In Pervasive and Mobile Computing, Elsevier, volume 38, 2017.

   BibTeX

   
   @article{georgievski2017:pmc, author = {I. Georgievski and T. A. Nguyen and F. Nizamic and B. Setz and A. Lazovik and M. Aiello}, title = {Planning meets activity recognition: Service coordination for intelligent buildings}, journal = {Pervasive and Mobile Computing}, year = {2017}, month = {July}, publisher = {Elsevier}, volume={38}, number={1}, pages={110-139}, url = {https://doi.org/10.1016/j.pmcj.2017.02.008}, }

   ---

   url
4. Metrics for Sustainable Data Centers (D. Vemula, B. Setz, G. S. VRK Rao, G. R. Gangadharan and M. Aiello), In IEEE Transactions on Sustainable Computing, 2017.

   BibTeX

   
   @article{Setz2017:sc, author = {D. Vemula and B. Setz and G. S. VRK Rao and G. R. Gangadharan and M. Aiello}, title = {Metrics for Sustainable Data Centers}, journal = {IEEE Transactions on Sustainable Computing}, year = {2017}, note = {To appear}, }

   ---

5. Power Management of Personal Computers based on User Behaviour (B. Setz, F. Nizamic, A. Lazovik and M. Aiello), In International Conference on Smart Cities and Green ICT Systems, 2016.

   BibTeX

   
   @inproceedings{setz2016:power, title = {Power Management of Personal Computers based on User Behaviour}, author = {B. Setz and F. Nizamic and A. Lazovik and M. Aiello}, year = {2016}, booktitle = {International Conference on Smart Cities and Green ICT Systems}, pages = {409--416}, url = {http://dx.doi.org/10.5220/0005762604090416}, }

   ---

   url

(For more publications go to Brian's publication page)