A Study on the Scalability of Hybrid LS-DYNA on Multicore Architectures
In this paper, the effects of differences in problem size, number of cores per processor, and interconnect on the Hybrid LS-DYNA’s performance are studied. The result shows that combination of these three factors determines when Hybrid LS-DYNA has a performance advantage over the MPP LS-DYNA.
ComputingTechnology-4.pdf
— 161.2 KB