Mapping of DNA sequencing and genetic algorithms to a multiprocessor (multicore) systems based in FPGAs: metodological aspects
| Grantee
| Jose Edinson Aedo Cobo |
| Date begin
| 10.2007 |
| Date end
| 07.2008 |
| Tutor
| Prof. Ivan Defilippis |
| Description
| In this research stay were defined two projects to develop considering a long term period. The first one was an academic project consisting in the design of a specialized course with emphasis in 32 bits processor and multicore architectures. The second was a research project called: "Mapping of DNA sequencing and genetic algorithms to a multiprocessor (multicore) system based in FPGAs". In this research project it is explored the design of multiprocessor architectures (multicore), using Microblaze and PowerPC processors, for executing efficiently genetic algorithms and DNA sequencing algorithms. It was used, the Xilinx Virtex II Pro FPGA as experimental platform. In this work, we analyze and define hardware architectural organizations with customized links between the processors and the memory in order to optimize the execution of both algorithms. |
| Objectives
| Analyze and explore the use of FPGAS multicore architectures for the efficient execution of genetic algorithms and DNA sequencing algorithms, and to design an advance course on high performance embedded system design. |
| Result
| Two specific algorithms were chosen as workload in this project: a genetic algorithm (GA) for multi-objective optimization and Smith-Waterman (SW) algorithm for ADN local alignment. These algorithms were compiled using the PowerPC tool chain. Also instrumental codes using the internal timers were introduced in both algorithms to performance measure and profiling. It was built a reference architecture based on one PowerPC core. Then some heterogeneous multicore architectures were building with two core and four cores in order to test them with diverse algorithm partitions. For this exploration, it was used Xilinx EDK 8.1 tool and Xilinx Virtex II FPGA development system. Finally, concerning to the academic project, it was make the complete design of the advance course. The laboratory component was designed considering the FPGA used in the research project. |
|
