Skip to main content

Advances, Systems and Applications

Table 10 Accumulated exclusive time of processes of StarStream in case of 16 threads

From: Investigating performance metrics for container-based HPC environments using x86 and OpenPOWER systems

  

x86

OpenPOWER

Processes

Interpretation

Exclusive Execution Time(Sec)

% of Total Execution Time

Exclusive Execution Time(Sec)

% of Total Execution Time

tuned_STREAM_Add._omp_fn.5:stream.inst.c

OpenMP implementation of Stream add kernel

224.717

25.29

-

-

tuned_STREAM_Triad._omp_fn.6:stream.inst.c

OpenMP implementation of Stream triad kernel

224.553

25.27

-

-

tuned_STREAM_Copy._omp_fn.3:stream.inst.c

OpenMP implementation of Stream copy kernel

169.924

19.12

-

-

tuned_STREAM_Scale._omp_fn.4:stream.inst.c

OpenMP implementation of Stream scale kernel

169.681

19.09

-

-

HPCC_Stream._omp_fn.1:stream.inst.c

HPCC parent stream function

056.604

06.37

601.192

59.78

Unknown

not reported

-

-

170.243

16.93

MPIDI_CH3I_Progress

maintain process progress state for asynchronous communications

-

-

100.768

10.02

MPID_nem_tcp_connpoll

tcp communication between processes

-

-

085.760

08.53

computeSTREAMerrors

compute the average errors for each array

020.056

02.26

020.940

02.08