Parallelization benchmarks of plane wave self-consistent field calculations on an IBM BlueGene supercomputer
Abstract
A plane wave self-consistent field calculation was parallelized over two node cards of an IBM BlueGene/P supercomputer. A maximum speedup of 9× was measured in Virtual Node execution mode. The observed performance improvements are consistent with Amdahl’s law for an algorithm that is approximately 89% parallelizable.
Downloads
Published
2016-08-18
Issue
Section
Poster Session PB
How to Cite
[1]
“Parallelization benchmarks of plane wave self-consistent field calculations on an IBM BlueGene supercomputer”, Proc. SPP, vol. 34, no. 1, p. SPP-2016-PB-19, Aug. 2016, Accessed: Apr. 16, 2026. [Online]. Available: https://proceedings.spp-online.org/article/view/SPP-2016-PB-19








