| Abstract: |
|
Understanding the measured performance of parallel applications in real systems is difficult -- with the aim to utilize the resources available, optimizations deployed in hardware and software layers include complex schemes. However, in order to identify a bottleneck the performance must be assessed.
This paper introduces PIOsimHD, an event-driven simulator for MPI-I/O applications and the underlying (heterogeneous) cluster computers. With the help of the simulator runs of MPI-I/O applications can be conducted in-silico; this includes detailed simulation of collective communication patterns as well as simulation of parallel I/O. The simulation provides an upper bounds for expected performance and helps to assess observed performance.
Together with HDTrace, an environment which allows to trace the behavior of MPI programs and internals of MPI and PVFS, PIOsimHD enables to localize inefficiencies, to conduct research on optimizations for communication algorithms, and to evaluate arbitrary and future systems.
In this paper the simulator is introduced and an excerpt of the conducted validation is presented, which demonstrates the surprising accuracy of the models for our cluster. |
|
