Michael Lescisin
ABSTRACT
A computer program, at its most basic level is a series of low level processor instructions which are executed sequentially. These instructions take time to execute, thus longer programs have longer execution times. One way to decrease the execution time for a program is to decrease the required time for each instruction. This is called frequency scaling. The disadvantage of frequency scaling is that running a processor at higher speeds causes it to generate more heat and consume more power. The solution to the problem of frequency scaling is to, instead of decreasing the time to execute an instruction, increase the number of instructions that can be run in a given amount of time, by running these instructions in parallel. In this paper we present a solution for using many off-the-shelf computers to build a computing cluster which will accelerate computing performance by running tasks in parallel.
- D. Thomasset and M. Grobe, "An introduction to the Message Passing Interface (MPI) using C", http://condor.cc.ku.edu/~grobe/docs/intro-MPI-C.shtml, {Online; accessed January 2016}.Google Scholar
- P. Abrahamsson, S. Helmer, N. Phaphoom, L. Nicolodi, N. Preda, L. Miori, M. Angriman, J. Rikkia, X. Wang, K. Hamily, and S. Bugoloni, "Affordable and Energy-Efficient Cloud Computing Clusters: The Bolzano Raspberry Pi Cloud Cluster Experiment", 5th IEEE International Conference on Cloud Computing Technology and Science (CloudCom 2013), Bristol, UK, pp. 170--175. Google ScholarDigital Library
Recommendations
Exploring Processor Design Options for Java-Based Middleware
ICPP '05: Proceedings of the 2005 International Conference on Parallel ProcessingJava-based middleware is a rapidly growing workload for high-end server processors, particularly Chip Multiprocessors (CMP). To help architects design future microprocessors to run this important new workload, we provide a detailed characterization of ...
Towards architecture-level middleware-enabled exception handling of component-based systems
CBSE '11: Proceedings of the 14th international ACM Sigsoft symposium on Component based software engineeringException handling is a practical and important way to improve the availability and reliability of a component-based system. The classical code-level exception handling approach is usually applied to the inside of a component, while some exceptions can ...
Power-efficient computing for compute-intensive GPGPU applications
PACT '12: Proceedings of the 21st international conference on Parallel architectures and compilation techniquesThe peak performance of graphics processing units (GPUs) has traditionally been increased by increasing the number of compute resources and/or their frequency. However, these approaches significantly increase the power consumption of GPUs. Consequently, ...
Comments