Visible to Intel only — GUID: GUID-F5973030-DFA8-405C-AE09-1F96CDEB8951
Measuring Communication and Computation Overlap
Measuring Pure Communication Time
Iallgather
Iallgather_pure
Iallgatherv
Iallgatherv_pure
Iallreduce
Iallreduce_pure
Ialltoall
Ialltoall_pure
Ialltoallv
Ialltoallv_pure
Ibarrier
Ibarrier_pure
Ibcast
Ibcast_pure
Igather
Igather_pure
Igatherv
Igatherv_pure
Ireduce
Ireduce_pure
Ireduce_scatter
Ireduce_scatter_pure
Iscatter
Iscatter_pure
Iscatterv
Iscatterv_pure
Visible to Intel only — GUID: GUID-F5973030-DFA8-405C-AE09-1F96CDEB8951
Benchmark Methodology
This section describes:
Different ways to manage Intel(R) MPI Benchmarks control flow
Command-line syntax for running the benchmarks
Sample output data from Intel(R) MPI Benchmarks
- Control Flow
- Command-line Control
- Benchmark Selection Arguments
- -npmin Option
- -multi Option
- -off_cache cache_size[,cache_line_size] Option
- -iter Option
- -iter_policy Option
- -time Option
- -mem Option
- -input <File> Option
- -msglen <File> Option
- -map PxQ Option
- -include [[benchmark1] benchmark2 …]
- -exclude [[benchmark1] benchmark2 …]
- -msglog [<minlog>:]<maxlog>
- -thread_level Option
- -sync Option
- -imb_barrier Option
- -root_shift Option
- -data_type Option
- -red_data_type Option
- -contig_type Option
- -zero_size Option
- Command-Line Control for IMB-P2P Benchmarks
- Command-Line Control for IMB-MT Benchmarks
- Parameters Controlling Intel(R) MPI Benchmarks
- Hard-Coded Settings
- Checking Results
- Output