Skip to main content
. 2008 Jul 21;9:315. doi: 10.1186/1471-2105-9-315

Figure 1.

Figure 1

Observed and predicted run times of the EPISNPmpi program on Minnesota Supercomputing Institute's 2.6 GHz IBM BladeCenter Linux cluster (Blade) and the SGI Altix XE 1300 Linux cluster system with 2.66 GHz Intel Clovertown processor (Calhoun). The observed run times (circles representing Blade and squares representing Calhoun) matched well with the predicted run times under ideal speedup and scalability (solid line representing Blade and dotted line representing Calhoun). Analyses in this figure used a hypothetical GWAS data set with 50,000 SNPs and 2000 individuals.