Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
[prof] in gg_tt.mad, profile ranmar (in ranmar.f: but this causes dou…
…ble counting in x2f for instance) - small contribution, will revert The contribution is small because it does not make Fortran Other decrease... (while x2f increases due to profiling overhead) ./build.none_d_inl0_hrd0/madevent_cpp < /tmp/avalassi/input_ggtt_x1_cudacpp [COUNTERS] PROGRAM TOTAL : 0.7472s [COUNTERS] Fortran Other ( 0 ) : 0.0105s [COUNTERS] Fortran X2F ( 1 ) : 0.0212s for 16399 events => throughput is 1.29E-06 events/s [COUNTERS] Fortran PDF ( 2 ) : 0.0943s for 32768 events => throughput is 2.88E-06 events/s [COUNTERS] Fortran final_I/O ( 3 ) : 0.2576s for 16399 events => throughput is 1.57E-05 events/s [COUNTERS] CudaCpp HEL ( 5 ) : 0.0007s [COUNTERS] CudaCpp MEs ( 6 ) : 0.0860s for 16384 events => throughput is 5.25E-06 events/s [COUNTERS] Fortran initial_I/O ( 7 ) : 0.0638s [COUNTERS] Fortran ranmar ( 12 ) : 0.0057s for 114719 events => throughput is 4.93E-08 events/s [COUNTERS] Fortran scale ( 13 ) : 0.0098s for 16384 events => throughput is 6.00E-07 events/s [COUNTERS] Fortran rewgt ( 14 ) : 0.0508s for 16384 events => throughput is 3.10E-06 events/s [COUNTERS] Fortran unwgt ( 16 ) : 0.1470s for 16384 events => throughput is 8.97E-06 events/s [COUNTERS] PROGRAM initial_I/O ( 19 ) : 0.0638s [COUNTERS] PROGRAM sample_full ( 20 ) : 0.6805s
- Loading branch information