Porting Autodock to CUDA
2010pp. 1–8
Citations Over TimeTop 10% of 2010 papers
Abstract
This paper is a report on the migration of the molecular docking application, “Autodock” to NVIDIA CUDA. Autodock is a Drug Discovery Tool that uses a Genetic Algorithm to find the optimal docking position of a ligand to a protein. Speedup of Autodock greatly benefits the drug discovery process. In this paper, we show how significant speed up of Autodock can be achieved using NVIDIA CUDA. This paper describes the strategy of porting the Genetic Algorithm to CUDA. Three different parallel design alternatives are discussed. The resultant implementation features ~50x speedup on the fitness function evaluation and 10x to 47x speedup on the core genetic algorithm.
Related Papers
- → Implementation of a Non-bonded Interaction Calculation Algorithm for the Cell Architecture(2009)4 cited
- → Speeding up model building for ECGA on CUDA platform(2013)2 cited
- → Innovative approach for porting existing CPU program to its CUDA program(2015)
- → CUDA GPU Programming Applied to HSI Exploitation(2017)
- → CUDA-Parttree: A Multiple Sequence Alignment Parallel Strategy in GPU(2019)