Abstract
We have developed a high performance 3D convolution library for Protein Docking on IBM Blue Gene. The algorithm is designed to exploit slight locality of memory access in 3D-FFT by making full use of a cache memory structure. The ID-FFT used in the 3D convolution is optimized for PowerPC 440 FP2 processors. The number of SIMOMD instructions is minimized by simultaneous computation of two ID-FFTs. The high performance 3D convolution library achieves up to 2.16 Gflops (38.6% of peak) per node. The total performance of a shape complementarity search is estimated at 7 Tflops with the 4-rack Blue Gene system (4096 nodes). © Springer-Verlag Berlin Heidelberg 2007.
Cite
CITATION STYLE
Nukada, A., Hourai, Y., Nishida, A., & Akiyama, Y. (2007). High performance 3D convolution for protein docking on IBM blue gene. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4742 LNCS, pp. 958–969). Springer Verlag. https://doi.org/10.1007/978-3-540-74742-0_84
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
