OV - Compare Index

Help is available by moving the cursor above any symbol or by checking MAQAO website.

▶Compared Reports

r0: avbp-06may_100iter_m36_ov1_gcc
r1: avbp-06may_100iter_m36_ov2

Global Metrics

Metric		r0	r1
Total Time (s)		450.32	342.29
Max (Thread Active Time) (s)		419.94	336.05
Average Active Time (s)		394.24	335.66
Activity Ratio (%)		93.9	99.9
Average number of active threads		33.797	35.958
Affinity Stability (%)		95.6	100
Time in analyzed loops (%)		67.6	72.9
Time in analyzed innermost loops (%)		54.5	49.4
Time in user code (%)		66.8	71.0
Compilation Options Score (%)		37.5	100
Array Access Efficiency (%)		66.6	80.8

Potential Speedups
Iterations Count		Not Available	1.08
Perfect Flow Complexity		1.01	1.01
Perfect OpenMP + MPI + Pthread		1.00	1.04
Perfect OpenMP + MPI + Pthread + Perfect Load Distribution		1.07	1.13
No Scalar Integer	Potential Speedup	1.06	1.21
No Scalar Integer	Nb Loops to get 80%	21	28
FP Vectorised	Potential Speedup	1.23	1.15
FP Vectorised	Nb Loops to get 80%	34	38
Fully Vectorised	Potential Speedup	2.14	2.03
Fully Vectorised	Nb Loops to get 80%	41	41
Only FP Arithmetic	Potential Speedup	1.20	1.61
Only FP Arithmetic	Nb Loops to get 80%	36	41
Data In L1 Cache	Potential Speedup	Not Available	1.01
Data In L1 Cache	Nb Loops to get 80%	Not Available	3

Cumulated Speedup If No Scalar Integer

Cumulated Speedup If FP Vectorized

Cumulated Speedup If Fully Vectorized

Cumulated Speedup If Only FP Arithmetic

Cumulated Speedup if Data in L1 cache

Loop Based Profiles

Innermost / Single Loops

Inbetween Loops

Outermost Loops

Cumulated Coverage With All Loops

Innermost Loop Based Profiles

Coverage

Count

Application Categorization

Time

Coverage

Compilation Options

Source Object	Issue
▼AVBP_V7.16.0.KRAKENGNU–
▼cons_tens.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼msource_cell.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼efcy_dyn.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼bndflux_les2.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼gradqen.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼grad_4obj.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼mass_product.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼euler_timestep.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼scatter_o_add.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼avis_lp_rre.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼specsource_cell.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼mod_pmesh_transfer.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼update.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼heatflux_nv2.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼central_nv_bnd_generic.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼temperature.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼mod_pmesh_scatter_add.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼div.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼FE_add_dw.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼wtowp.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼savis_Colin_NS.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼avis_lp.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼scatter_grad.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼scatter_o_sub.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼get_bcells.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼scatter_add.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼rot_2delta.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼specflux_visc_c_nv.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼rrate_cell.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼compute_FE_implicit_residual.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼get_uvwT.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼wale_cell.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼specflux_invc.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼gather_o_cpy.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼nsflux_les.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼dlongc.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼thermo_variables.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼prebound.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼laxwe.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼central_nv.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼scheme.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼ave.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼scale.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼boxe_2delta.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼calc_diffus.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼get_Y.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼stress_nv2.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼eflux.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼mod_adj_graph.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼central.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼mod_copy.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼savis_spec.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼calc_visc_eff.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼cons_tens_cell.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼square.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.
▼compute_diffus_max.f90–
○	-O3 or -Ofast is missing.
○	-march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native.
○	-funroll-loops is missing.

Source Object	Issue
▼AVBP_V7.16.0.KRAKEN–
▼cons_tens.f90–
○
▼msource_cell.f90–
○
▼scatter_o_add.f90–
○
▼scale.f90–
○
▼bndflux_les2.f90–
○
▼efcy_dyn.f90–
○
▼mod_pmesh_scatter_add.f90–
○
▼nsflux_les.f90–
○
▼update_rho.f90–
○
▼ns_timestep.f90–
○
▼euler_timestep.f90–
○
▼scheme.f90–
○
▼mod_adj_graph.f90–
○
▼specsource_cell.f90–
○
▼wtowp.f90–
○
▼get_Y.f90–
○
▼heatflux_nv2.f90–
○
▼mod_qsort.f90–
○
▼temperature.f90–
○
▼grad_4obj.f90–
○
▼avis_lp.f90–
○
▼FE_add_dw.f90–
○
▼specflux_invc.f90–
○
▼mod_pmesh_transfer.f90–
○
▼compute_FE_implicit_residual.f90–
○
▼scatter_grad.f90–
○
▼div.f90–
○
▼get_bcells.f90–
○
▼update.f90–
○
▼rot_2delta.f90–
○
▼specflux_visc_c_nv.f90–
○
▼rrate_cell.f90–
○
▼mass_product.f90–
○
▼scatter_o_sub.f90–
○
▼wale_cell.f90–
○
▼scatter_add.f90–
○
▼gather_o_cpy.f90–
○
▼avis_lp_rre.f90–
○
▼mod_bsearch.f90–
○
▼thermo_variables.f90–
○
▼prebound.f90–
○
▼stress_nv2.f90–
○
▼central_nv.f90–
○
▼laxwe.f90–
○
▼savis_Colin_NS.f90–
○
▼calc_visc_eff.f90–
○
▼boxe_2delta.f90–
○
▼calc_diffus.f90–
○
▼central.f90–
○
▼central_nv_bnd_generic.f90–
○
▼get_uvwT.f90–
○
▼gradqen.f90–
○
▼ave.f90–
○
▼mod_copy.f90–
○
▼savis_spec.f90–
○
▼velocity_group.f90–
○
▼cons_tens_cell.f90–
○
▼eflux.f90–
○
▼compute_diffus_max.f90–
○

Path Count Profiles

Coverage

Count

Low Iteration Count Profiles

Coverage

Count

Average Number of Active Threads

Run 1 - avbp-06may_100iter_m36_ov1_gcc

Run 2 - avbp-06may_100iter_m36_ov2

Experiment Summaries

	r0	r1
Experiment Name
Application	/scratch/exter/camus/avbp/avbp-7.16.0/HOST/KRAKENGNU/BIN/AVBP_V7.16.0.KRAKENGNU	/scratch/exter/camus/avbp/avbp-7.16.0/HOST/KRAKEN/BIN/AVBP_V7.16.0.KRAKEN
Timestamp	2025-05-13 16:34:55	2025-05-12 11:59:26
Experiment Type	MPI;	same as r0
Machine	node186	node025
Architecture	x86_64	same as r0
Micro Architecture	SKYLAKE	same as r0
Model Name	Intel(R) Xeon(R) Gold 6140 CPU @ 2.30GHz	same as r0
Cache Size	25344 KB	same as r0
Number of Cores	18	same as r0
Maximal Frequency	3.7 GHz	same as r0
OS Version	Linux 4.18.0-553.el8_10.x86_64 #1 SMP Fri May 24 13:05:10 UTC 2024	same as r0
Architecture used during static analysis	x86_64	same as r0
Micro Architecture used during static analysis	SKYLAKE	same as r0
Compilation Options	AVBP_V7.16.0.KRAKENGNU: GNU Fortran2008 11.2.0 -mtune=generic -march=x86-64 -g -O2 -fbacktrace -ffree-line-length-none -fconvert=big-endian -fallow-argument-mismatch -fintrinsic-modules-path /home/logiciels/local/gcc/11.2.0/bin/../lib/gcc/x86_64-pc-linux-gnu/11.2.0/finclude -fpre-include=/usr/include/finclude/math-vector-fortran.h	AVBP_V7.16.0.KRAKEN: Intel(R) Fortran Intel(R) 64 Compiler Classic for applications running on Intel(R) 64, Version 2021.10.0 Build 20230609_000000 -I/softs/local_intel/phdf5/1.8.20/include -I/softs/local_intel/parmetis/403_64/include -I/softs/local_intel/ptscotch/6.0.5a/include -I. -I../SOURCES/GENERIC/ -IAMR_INTERFACE/ -IBNDY/ -ICFD/ -ICHEM/ -ICHEM/ANALYTIC/ -ICHEM/ANALYTIC/LIB/ -ICHEM/HYB/ -ICHEM/NOX/ -ICHEM/SOOT_ANALYTIC/ -ICOMMON/ -ICONDUCTION/ -ICOUPLING/ -IGENERIC/ -IIO/ -ILAGRANGE/ -ILAGRANGE/SOOT_EL/ -ILES/ -IMAIN/ -IMAIN/COMPUTE/ -IMAIN/SLAVE/ -INUMERICS/ -IPARSER/ -IPLASMA/ -IPLASMA/CHEMISTRY/ -IPLASMA/CHEMISTRY/CUSTOM_KINETICS_LIB/ -IPLASMA/DRIFTDIFFUSION/ -IPLASMA/DRIFTDIFFUSION/SCHEMES/ -IPLASMA/ELECTROMAG/ -IPLASMA/EULER/ -IPLASMA/FREEZE/ -IPLASMA/PHOTO/ -IPLASMA/THERMO/ -IPMESH/generic/ -IPMESH/interf_avbp/ -IPMESH/interp_tree_search/ -IPMESH/pmeshlib/ -IPMESH/pproc/ -ISMOOTH/ -ITTC/ -ITTC/LES/ -I/softs/intel/oneapi/mpi/2021.10.0//include -I/softs/intel/oneapi/mpi/2021.10.0/include -g -O3 -fpp -traceback -fno-alias -ip -assume byterecl -convert big_endian -align -march=core-avx2 -fma -axCORE-AVX2 -DHAS_PMETIS -DPARMETIS4 -DMETIS5 -DHAS_PTSCOTCH -c -o GENERIC/gather_o_cpy.o
Number of processes observed	36	same as r0
Number of threads observed	36	same as r0
Frequency Driver	intel_cpufreq	same as r0
Frequency Governor	performance	same as r0
Huge Pages	always	same as r0
Hyperthreading	off	same as r0
Number of sockets	2	same as r0
Number of cores per socket	18	same as r0
MAQAO version	2.21.4	same as r0
MAQAO build	d26a8e0cc0244e3e28a60368aba55c6c7d2cdaba::20250506-180849	same as r0
Comments		same as r0

Report Configuration

▶Compared Reports

Global Metrics

Cumulated Speedup If No Scalar Integer

Cumulated Speedup If FP Vectorized

Cumulated Speedup If Fully Vectorized

Cumulated Speedup If Only FP Arithmetic

Cumulated Speedup if Data in L1 cache

Loop Based Profiles

Innermost / Single Loops

Inbetween Loops

Outermost Loops

Cumulated Coverage With All Loops

Innermost Loop Based Profiles

Coverage

Count

Application Categorization

Time

Coverage

Compilation Options

Path Count Profiles

Coverage

Count

Low Iteration Count Profiles

Coverage

Count

Average Number of Active Threads

Run 1 - avbp-06may_100iter_m36_ov1_gcc

Run 2 - avbp-06may_100iter_m36_ov2

Experiment Summaries