CUDA RNAfold

Abstract

Spectra-Based Fault Localisation (SBFL) aims to assist de-bugging by applying risk evaluation formulae (sometimes called suspiciousness metrics) to program spectra and ranking statements according to the predicted risk. Designing a risk evaluation formula is often an intuitive process done by human software engineer. This paper presents a Genetic Programming approach for evolving risk assessment formulae. The empirical evaluation using 92 faults from four Unix utilities produces promising results 1. GP-evolved equations can consistently outperform many of the human-designed formulae, such as Tarantula, Ochiai, Jaccard, Ample, and Wong1/2, up to 5.9 times. More importantly, they can perform equally as well as Op2, which was recently proved to be optimal against If-Then-Else-2 (ITE2) structure, or even outperform it against other program structures. 1 The program spectra data used in the paper, as well as the complete empirical results, are available from:http://www.cs.ucl.ac.uk/staff/s.yoo/evolving-sbfl.html. UCL DEPARTMENT OF COMPUTER SCIENCE Research Note W. B. Langdon and Ronny Lorenz AbstractWe add CUDA GPU C program code to RNAfold to enable both it to be run on nVidia gaming graphics hardware and so that many thousands of RNA secondary structures can be computed in parallel. RNAfold predicts the folding pattern for RNA molecules by using O(n 3 ) dynamic programming matrices to minimise the free energy of treating them as a sequence of bases. We benchmark RNAfold on RNA STRAND and artificial sequences of upto 30 000 bases on two GPUs and a GPGPU Tesla. The speed up is variable but up to 14 times.

Evolving AVX512 Parallel C Code Using GP

“…the Intel Phi cards. (ViennaRNA also includes our GPU version of RNAfold [4].) Since then AVX has been made available in certain top end Intel CPU chips and so we renewed our attempt to automatically port RNAfold to AVX512 hardware.…”