With the rapid advancement of high throughput sequencing, large numbers of genetic markers can be readily and cheaply acquired, but most current software packages for genetic map construction cannot handle such dense input. Modern computer architectures and server farms represent untapped resources that can be used to enable higher marker densities to be processed in tractable time. Here we present a pipeline using a modified version of OneMap that parallelizes over bottleneck functions and achieves substantial speedups for producing a high density linkage map (N = 20,000). Using simulated data we show that the outcome is as accurate as the traditional pipeline. We further demonstrate that there is a direct relationship between the number of markers used and the level of deviation between true and estimated order, which in turn impacts the final size of a genetic map.