Abstract—Massively parallel computers have found significantinterest from researchers in recent years. These machines havebeen used for complex and sophisticated simulations, such as,the brain functions simulation. Due to the increase in thepower demand by massively parallel computers and it’s moveinto the exascale computing in future, temperature and powerconsumption has become a major constrains, many researchesaimed to address the power efficiency and temperature problems.In this work, we studied the effect of the programminglanguages statements on system power consumption for threecontrol loop statements. These statements are: For Loop, WhileLoop, and Do-While Loop.For each of the three statements six measurements wereobtained: average temperature for the control loop, averagemaximum reached temperature for the control loop, averagepercentage of execution time when the control loop reach themaximum temperature, average number of times the control loopreach the maximum temperature, average power consumption forthe control loop, and average execution time for the control loop.The findings indicated that the For loop statement was themost efficient control loop statement. Also, it showed that theWhile loop statement was the worst in efficiency.Keywords—High Performance Computing, Exascale Computing,Power-Efficiency, Power-Efficient Programming LanguageStatements, Temperature Monitoring.
Evaluation of Control Loop Statements Power Efficiency: An Experimental Study
2015-06-15
Proceedings of INFOS2014; 2014
Article (Journal)
Electronic Resource
English
DDC: | 629 |
Evaluation of Subjective Probability Statements
AIAA | 2010
|Evaluation of Subjective Probability Statements
British Library Conference Proceedings | 2010
|Online Contents | 1997
Elsevier | 2012