options

Functions and Loops

9 loops and 2 functions have been discarded from the report because their ratio ((Max Inclusive Time Over Threads * 100) / Max Thread Active Time) is lower than the threshold set by object_coverage_threshold (0.01%). It represents about 0.00% of the application. To include them, change the value of object_coverage_threshold in the experiment directory configuration file, then rerun the command with the additionnal parameter --force-static-analysis.
Inclusive metrics are only related to the given object code and do not include other external objects / libraries.

Columns Filter

Max Thread Time / Walltime run_0 (%) Coverage run_0 (%) Coverage Excluding Loops run_0 (%) Max Inclusive Time Over Threads run_0 (s) Max Exclusive Time Over Threads run_0 (s) Inclusive Time w.r.t. Wall Time run_0 (s) Exclusive Time w.r.t. Wall Time run_0 (s) Nb Threads run_0 Deviation (coverage) run_0 Deviation (walltime) run_0 Categories run_0 GFLOPS run_0 Compilation Options Max Thread Time / Walltime Coverage Coverage Excluding Loops Max Inclusive Time Over Threads Max Exclusive Time Over Threads Inclusive Time w.r.t. Wall Time Exclusive Time w.r.t. Wall Time Nb Threads Deviation (coverage) Deviation (walltime) Categories GFLOPS Compilation Options
NameModuleMax Thread Time / Walltime run_0 (%)Coverage run_0 (%)Coverage Excluding Loops run_0 (%)Max Inclusive Time Over Threads run_0 (s)Max Exclusive Time Over Threads run_0 (s)Inclusive Time w.r.t. Wall Time run_0 (s)Exclusive Time w.r.t. Wall Time run_0 (s)Nb Threads run_0Deviation (coverage) run_0Deviation (walltime) run_0Categories run_0GFLOPS run_0Compilation Options
main+attention-avx51299.4199.580.003.570.003.570.0010.000.00Exe (%): 100.009.70 --driver-mode=g++ --intel -O3 -g -fno-omit-frame-pointer -grecord-command-line -x GRANITERAPIDS -mprefer-vector-width=512 -qopt-report=2 -ffp-model=fast=2 attention.cpp -o attention-avx512 -fveclib=SVML -dumpdir attention-avx512-
Loop 23 - random.tcc:330-336 - attention-avx512 [...]0.000.000.000.000.000.000.0000.000.000.00
Loop 58 - attention.cpp:26-193 - attention-avx512 [...]+0.009.760.000.350.000.350.0000.000.000.00
Loop 59 - attention.cpp:26-193 - attention-avx512 [...]+0.009.760.000.350.000.350.0000.000.000.00
Loop 62 - attention.cpp:27-33 - attention-avx512+0.979.760.980.350.040.350.0410.000.0023.37
Loop 63 - attention.cpp:30-31 - attention-avx5128.778.798.790.310.310.310.3110.000.007.11
Loop 57 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 60 - attention.cpp:27-98 - attention-avx512 [...]+0.000.000.000.000.000.000.0000.000.000.00
Loop 61 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 65 - attention.cpp:26-171 - attention-avx512 [...]+0.009.620.000.350.000.350.0000.000.000.00
Loop 66 - attention.cpp:26-171 - attention-avx512 [...]+0.009.620.000.350.000.350.0000.000.000.00
Loop 69 - attention.cpp:27-33 - attention-avx512+0.979.620.980.350.040.350.0410.000.0023.06
Loop 70 - attention.cpp:30-31 - attention-avx5128.638.658.650.310.310.310.3110.000.007.26
Loop 64 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 67 - attention.cpp:27-98 - attention-avx512 [...]+0.000.000.000.000.000.000.0000.000.000.00
Loop 68 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 51 - attention.cpp:26-215 - attention-avx512 [...]+0.009.620.000.350.000.340.0000.000.000.00
Loop 52 - attention.cpp:26-215 - attention-avx512 [...]+0.009.620.000.350.000.340.0000.000.000.00
Loop 53 - attention.cpp:27-98 - attention-avx512 [...]+0.000.000.000.000.000.000.0000.000.000.00
Loop 54 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 55 - attention.cpp:27-33 - attention-avx512+0.429.620.420.350.020.340.0210.000.0053.10
Loop 50 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 56 - attention.cpp:30-31 - attention-avx5129.199.219.210.330.330.330.3310.000.006.86
Loop 72 - attention.cpp:156-156 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 38 - attention.cpp:26-262 - attention-avx512 [...]+0.0036.820.001.320.001.320.0000.000.000.00
Loop 39 - attention.cpp:26-262 - attention-avx512 [...]+0.0036.820.001.320.001.320.0000.000.000.00
Loop 42 - attention.cpp:27-33 - attention-avx512+5.0136.825.021.320.181.320.1810.000.0019.84
Loop 43 - attention.cpp:30-31 - attention-avx51231.7531.8031.801.141.141.141.1410.000.007.77
Loop 37 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 40 - attention.cpp:27-98 - attention-avx512 [...]+0.000.000.000.000.000.000.0000.000.000.00
Loop 41 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 32 - attention.cpp:43-284 - attention-avx512 [...]+0.003.070.000.110.000.110.0000.000.000.00
Loop 33 - attention.cpp:43-284 - attention-avx512 [...]+0.843.070.840.110.030.110.0310.000.0044.97
Loop 36 - attention.cpp:47-48 - attention-avx5121.811.811.810.060.060.060.0610.000.007.04
Loop 34 - attention.cpp:55-56 - attention-avx5120.420.420.420.020.020.020.0210.000.0010.67
Loop 35 - attention.cpp:52-53 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 31 - attention.cpp:47-48 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 71 - attention.cpp:157-160 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 45 - attention.cpp:98-241 - attention-avx512 [...]+0.000.000.000.000.000.000.0000.000.000.00
Loop 46 - attention.cpp:98-241 - attention-avx512 [...]+0.000.000.000.000.000.000.0000.000.000.00
Loop 47 - attention.cpp:239-241 - attention-avx512+0.000.000.000.000.000.000.0000.000.000.00
Loop 48 - attention.cpp:239-241 - attention-avx512+0.000.000.000.000.000.000.0000.000.000.00
Loop 44 - attention.cpp:240-241 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 49 - attention.cpp:240-241 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 25 - attention.cpp:26-306 - attention-avx512 [...]+0.0030.680.001.100.001.100.0000.000.000.00
Loop 26 - attention.cpp:26-306 - attention-avx512 [...]+0.0030.680.001.100.001.100.0000.000.000.00
Loop 27 - attention.cpp:27-98 - attention-avx512 [...]+0.000.000.000.000.000.000.0000.000.000.00
Loop 28 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
Loop 29 - attention.cpp:27-33 - attention-avx512+0.7030.680.701.100.031.100.0310.000.0033.60
Loop 30 - attention.cpp:30-31 - attention-avx51229.9429.9929.991.081.081.081.0810.000.009.39
Loop 24 - attention.cpp:30-31 - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
__svml_expf16_z0attention-avx5120.280.280.280.010.010.010.0110.000.00Math (%): 100.0028.20
__intel_avx_rep_memset+attention-avx5120.140.140.140.000.000.000.0010.000.00Memory (%): 100.0027.50
Loop 96 - - attention-avx5120.000.000.000.000.000.000.0000.000.000.00
×