원문정보
초록
영어
In order to speed up data processing in a signal monitoring and evaluation system, we need to use a parallel method. It is obvious that the traditional stand-alone store has no ability to satisfy the performance requirements, and the use of single core CPU is unable to content the severe requirement of speed. Consequently, multi-machine parallel acceleration technique based on MPI (cooperated with multi-core parallel acceleration technique based on OpenMP) can effectively solve all above problems. In this paper, a parallel distributed acceleration framework based on MPI and Open MP technology was given. Experimental tests were carried to verify our proposal. Finally, some suggestions to speed up the data processing was given.
목차
1. Introduction
2. Related Work
2.1. The Principles of MPI
2.2. The Principles of OpenMP
3. Experiment Principle
3.1. Serial Proposal
3.2. MPI & OpenMP Hybrid Parallel Proposal
3.3. The MPI Parallel Framework
3.4. Checking Data Parallelly
3.5. Reading Data Parallelly
3.6. Fitting Standard Deviation Parallelly
4. Test
4.1. Test Plan
4.2. Serial Model
4.3. Parallel Model
5. Conclusion
5.1. The Advantages of Parallel
5.2. Summary
References