abstract.tex

% -*- mode: latex; mode: visual-line; fill-column: 9999; coding: utf-8 -*-

% Max 200 words

\abstract[Summary]{%
The performance of biomolecular molecular dynamics simulations has steadily increased on modern high performance computing resources but acceleration of the analysis of the output trajectories has lagged behind so that analyzing simulations is becoming a bottleneck.
To close this gap, we studied the performance of parallel trajectory analysis with MPI and the Python \package{MDAnalysis} library on three different XSEDE supercomputers where trajectories were read from a Lustre parallel file system. 
Strong scaling performance was impeded by stragglers, MPI processes that were slower than the typical process. 
Stragglers were less prevalent for compute-bound workloads, thus pointing to file reading as a bottleneck for scaling.
However, a more complicated picture emerged in which both the computation and the data ingestion exhibited close to ideal strong scaling behavior whereas stragglers were primarily caused by either large MPI communication costs or long times to open the single shared trajectory file.
We improved overall strong scaling performance by either subfiling (splitting the trajectory into separate files) or MPI-IO with Parallel HDF5 trajectory files.
The parallel HDF5 approach resulted in near ideal strong scaling on up to 384 cores (16 nodes), thus reducing trajectory analysis times by two orders of magnitude compared to the serial approach.
}