Intel® MPI Benchmarks 3.2.4
This release includes the following updates as compared to the Intel® MPI Benchmarks 3.0:
New control flags
Better control of the overall repetition counts, run time, and memory exploitation
A facility to avoid cache re-usage of message buffers as far as possible
A fix of IMB-IO semantics
New benchmarks
Gather
Gatherv
Scatter
Scatterv
New command-line flags for better control
-off_cache
Use this flag when measuring performance on high speed interconnects or, in particular, across the shared memory within a node. Traditional Intel® MPI Benchmarks results included a very beneficial cache re-usage of message buffers which led to idealistic results. The flag -off_cache allows avoiding cache effects and lets the Intel® MPI Benchmarks use message buffers which are very likely not resident in cache.
-iter, -time
Use these flags for enhanced control of the overall run time, which is crucial for large clusters, where collectives tend to run extremely long in the traditional Intel® MPI Benchmarks settings.
CAUTION
In the Intel® MPI Benchmarks, the -time flag has been implemented as default.
-mem
Use this flag to determine an a priori maximum (per process) memory usage of the Intel® MPI Benchmarks for the overall message buffers.