Multi-Threaded DNA Tag/Anti-Tag Library Generator for Multi-Core Platforms
Abstract
Report Developed under CRADA 08-RI-CRADA-06. The report describes a new approach to the problem of generating DNA tag/anti-tag libraries used in experimental computing methods involving bio-molecules, and in biological assay methods. This approach couples multi-threaded coding methods and a highly parallel multi-population genetic algorithm to leverage performance gains made possible by the multi-core CPUs increasingly prevalent in today's commodity workstation computers. We explored and exploited algorithm and architecture trade-offs while developing a multi-threaded code that uses shared memory communication, and minimal synchronization between threads. We also describe experiments that evaluated performance and demonstrated ~ 5X-8X speedups on workstations with dual quad-core CPUs. We observe that coding effort using the C language and Pthreads parallel programming model is greatly reduced compared to two previous approaches that used the VHDL language run on reconfigurable hardware (FPGAs), and the C language with MPI API run on a cluster of computers.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 2009
- Accession Number
- ADA499723
Entities
People
- Arun Ravinfran
Organizations
- University of North Carolina at Charlotte