Abstract | ||
---|---|---|
In data center, the optimal server location selection is related to many factors, such as power, temperature and space location. In order to reduce the local hot point in computer room, decrease the energy consumption of cooling system and eliminate the power fragment of rack, this paper studies the optimization issue of server placement for the enterprise data center based on the joint perception of temperature and power. First, based on the deep learning algorithm, the model of the power, temperature and cooling power of equipment in computer room is established. The power curve model of the server is estimated by using the Conditional Generative Adversarial Network (CGAN), and then the optimal placement position of server is calculated by genetic algorithm. Through simulation experiments, it is proved that the method proposed in this paper is stable and feasible. Compared with the three placement algorithms, this method can reduce the cooling energy consumption by 4-6% and reduce the new power demand of 6-9%, which can effectively improve the energy consumption efficiency of data center. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1109/PADSW.2018.8644639 | 2018 IEEE 24th International Conference on Parallel and Distributed Systems (ICPADS) |
Keywords | Field | DocType |
Servers,Data centers,Cooling,Temperature distribution,Computational modeling,Energy consumption,Optimization | Rack,Computer science,Server,Real-time computing,Water cooling,Artificial intelligence,Deep learning,Energy consumption,Data center,Enterprise data management,Genetic algorithm | Conference |
ISSN | ISBN | Citations |
1521-9097 | 978-1-5386-7308-9 | 1 |
PageRank | References | Authors |
0.37 | 0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Longchuan Yan | 1 | 1 | 1.39 |
Wantao Liu | 2 | 73 | 8.29 |
Dongxia Bai | 3 | 21 | 3.94 |