diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/dgemm.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/dgemm.sirocco index 71459a7ff9fdac8958813af76131ba2ed214e450..213bfaf1c4350ce2311ec43e2206634de49a3d3e 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/dgemm.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/dgemm.sirocco @@ -7,12 +7,12 @@ # number of combinations 5 #################### -# COMB_4 +# COMB_2 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 0 #################### # DEV_0 @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb4) +# Model for cpu0_impl0 (Comb2) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,11 +36,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -744ff412 22118400 0.000000e+00 7.614739e+04 1.570335e+04 4.926736e+07 3.911128e+12 647 -b63efa5a 153600 0.000000e+00 5.080183e+01 9.743801e+00 2.861616e+06 1.507233e+08 56329 -9427311d 4646400 0.000000e+00 8.137654e+03 1.863872e+03 1.657640e+07 1.419696e+11 2037 -dc7b85e2 49766400 0.000000e+00 2.267816e+05 3.915016e+04 1.598810e+08 3.733864e+13 705 f74a4f19 88473600 0.000000e+00 4.662963e+05 5.310256e+04 3.273400e+08 1.546170e+14 702 +dc7b85e2 49766400 0.000000e+00 2.267816e+05 3.915016e+04 1.598810e+08 3.733864e+13 705 +9427311d 4646400 0.000000e+00 8.137654e+03 1.863872e+03 1.657640e+07 1.419696e+11 2037 +b63efa5a 153600 0.000000e+00 5.080183e+01 9.743801e+00 2.861616e+06 1.507233e+08 56329 +744ff412 22118400 0.000000e+00 9.214389e+04 2.022196e+04 1.782984e+08 1.722038e+13 1935 #################### # COMB_1 @@ -48,7 +48,7 @@ f74a4f19 88473600 0.000000e+00 4.662963e+05 5.310256e+04 3.27340 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -72,19 +72,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -744ff412 22118400 0.000000e+00 1.715310e+03 9.993558e+01 9.566285e+06 1.646484e+10 5577 -b63efa5a 153600 0.000000e+00 5.113274e+01 1.134809e+01 1.518642e+04 8.147708e+05 297 -9427311d 4646400 0.000000e+00 2.017727e+02 2.861061e+01 1.652922e+06 3.402201e+08 8192 -dc7b85e2 49766400 0.000000e+00 6.191028e+03 2.666186e+02 2.317302e+07 1.437309e+11 3743 f74a4f19 88473600 0.000000e+00 1.292684e+04 4.908603e+02 4.657542e+07 6.029413e+11 3603 +dc7b85e2 49766400 0.000000e+00 6.191028e+03 2.666186e+02 2.317302e+07 1.437309e+11 3743 +9427311d 4646400 0.000000e+00 2.017727e+02 2.861061e+01 1.652922e+06 3.402201e+08 8192 +b63efa5a 153600 0.000000e+00 5.113274e+01 1.134809e+01 1.518642e+04 8.147708e+05 297 +744ff412 22118400 0.000000e+00 1.678769e+03 1.038248e+02 6.356153e+07 1.071132e+11 37862 #################### -# COMB_2 +# COMB_0 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -98,7 +98,7 @@ f74a4f19 88473600 0.000000e+00 1.292684e+04 4.908603e+02 4.65754 # number of implementations 1 ##### -# Model for cuda0_impl0 (Comb2) +# Model for cuda0_impl0 (Comb0) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -108,19 +108,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -744ff412 22118400 0.000000e+00 1.694044e+03 9.446026e+01 9.451069e+06 1.606030e+10 5579 -b63efa5a 153600 0.000000e+00 6.271543e+01 1.282513e+01 1.994351e+04 1.303072e+06 318 -9427311d 4646400 0.000000e+00 2.001219e+02 2.795923e+01 1.640199e+06 3.346467e+08 8196 -dc7b85e2 49766400 0.000000e+00 6.117615e+03 2.941760e+02 2.339376e+07 1.434449e+11 3824 f74a4f19 88473600 0.000000e+00 1.279991e+04 4.666369e+02 4.609249e+07 5.907639e+11 3601 +dc7b85e2 49766400 0.000000e+00 6.117615e+03 2.941760e+02 2.339376e+07 1.434449e+11 3824 +9427311d 4646400 0.000000e+00 2.001219e+02 2.795923e+01 1.640199e+06 3.346467e+08 8196 +b63efa5a 153600 0.000000e+00 6.271543e+01 1.282513e+01 1.994351e+04 1.303072e+06 318 +744ff412 22118400 0.000000e+00 1.697238e+03 1.032817e+02 6.385010e+07 1.087701e+11 37620 #################### -# COMB_0 +# COMB_3 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -134,7 +134,7 @@ f74a4f19 88473600 0.000000e+00 1.279991e+04 4.666369e+02 4.60924 # number of implementations 1 ##### -# Model for cuda2_impl0 (Comb0) +# Model for cuda2_impl0 (Comb3) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -144,19 +144,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -744ff412 22118400 0.000000e+00 1.713334e+03 9.294880e+01 9.365083e+06 1.609274e+10 5466 -b63efa5a 153600 0.000000e+00 4.994209e+01 9.498420e+00 1.588158e+04 8.218495e+05 318 -9427311d 4646400 0.000000e+00 2.026661e+02 2.758349e+01 1.561340e+06 3.222922e+08 7704 -dc7b85e2 49766400 0.000000e+00 6.177427e+03 2.260170e+02 2.327037e+07 1.439434e+11 3767 f74a4f19 88473600 0.000000e+00 1.286952e+04 4.739314e+02 4.932886e+07 6.356995e+11 3833 +dc7b85e2 49766400 0.000000e+00 6.177427e+03 2.260170e+02 2.327037e+07 1.439434e+11 3767 +9427311d 4646400 0.000000e+00 2.026661e+02 2.758349e+01 1.561340e+06 3.222922e+08 7704 +b63efa5a 153600 0.000000e+00 4.994209e+01 9.498420e+00 1.588158e+04 8.218495e+05 318 +744ff412 22118400 0.000000e+00 1.713334e+03 9.294880e+01 9.365083e+06 1.609274e+10 5466 #################### -# COMB_3 +# COMB_4 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -170,7 +170,7 @@ f74a4f19 88473600 0.000000e+00 1.286952e+04 4.739314e+02 4.93288 # number of implementations 1 ##### -# Model for cuda3_impl0 (Comb3) +# Model for cuda3_impl0 (Comb4) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -180,9 +180,9 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -744ff412 22118400 0.000000e+00 1.680328e+03 1.055142e+02 9.028404e+06 1.523050e+10 5373 -b63efa5a 153600 0.000000e+00 5.000193e+01 1.025088e+01 7.350283e+03 3.829751e+05 147 -9427311d 4646400 0.000000e+00 1.988673e+02 2.685130e+01 1.504233e+06 3.045963e+08 7564 -dc7b85e2 49766400 0.000000e+00 6.051978e+03 2.759693e+02 2.305804e+07 1.398369e+11 3810 f74a4f19 88473600 0.000000e+00 1.258002e+04 4.620060e+02 4.874757e+07 6.140725e+11 3875 +dc7b85e2 49766400 0.000000e+00 6.051978e+03 2.759693e+02 2.305804e+07 1.398369e+11 3810 +9427311d 4646400 0.000000e+00 1.988673e+02 2.685130e+01 1.504233e+06 3.045963e+08 7564 +b63efa5a 153600 0.000000e+00 5.000193e+01 1.025088e+01 7.350283e+03 3.829751e+05 147 +744ff412 22118400 0.000000e+00 1.680328e+03 1.055142e+02 9.028404e+06 1.523050e+10 5373 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/dlauum.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/dlauum.sirocco new file mode 100644 index 0000000000000000000000000000000000000000..732dbabb74cab36d2231eeb1b8c849a583538e76 --- /dev/null +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/dlauum.sirocco @@ -0,0 +1,40 @@ +################## +# Performance Model Version +45 + +#################### +# COMBs +# number of combinations +1 +#################### +# COMB_2 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +0 +#################### +# DEV_0 +# device id +0 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cpu0_impl0 (Comb2) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +901ddf46 7372800 0.000000e+00 1.649855e+04 3.052838e+03 3.877159e+06 6.615766e+10 235 + diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/dplgsy.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/dplgsy.sirocco index e63e5c40cda42ba61ad702603be4fe23cd8a1fa1..0ba64725206b6de086b7024075292c5ca8f18ada 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/dplgsy.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/dplgsy.sirocco @@ -7,12 +7,12 @@ # number of combinations 1 #################### -# COMB_4 +# COMB_2 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 0 #################### # DEV_0 @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb4) +# Model for cpu0_impl0 (Comb2) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,9 +36,9 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -901ddf46 7372800 0.000000e+00 9.359014e+03 4.860060e+02 2.358472e+06 2.213249e+10 252 -5c532f3b 51200 0.000000e+00 9.876737e+01 1.343538e+01 1.590451e+06 1.599914e+08 16103 -69439308 1548800 0.000000e+00 2.134161e+03 1.815312e+02 4.673812e+06 1.004683e+10 2190 -d48836c9 16588800 0.000000e+00 2.092160e+04 1.229935e+03 3.661280e+06 7.686456e+10 175 a0677317 29491200 0.000000e+00 3.676615e+04 1.578675e+03 5.110494e+06 1.882396e+11 139 +d48836c9 16588800 0.000000e+00 2.092160e+04 1.229935e+03 3.661280e+06 7.686456e+10 175 +69439308 1548800 0.000000e+00 2.134161e+03 1.815312e+02 4.673812e+06 1.004683e+10 2190 +5c532f3b 51200 0.000000e+00 9.876737e+01 1.343538e+01 1.590451e+06 1.599914e+08 16103 +901ddf46 7372800 0.000000e+00 1.139975e+04 2.282146e+03 2.380267e+07 2.822191e+11 2088 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/dpotrf.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/dpotrf.sirocco index d84aa49cd289ce74ec56b29c64beca314131a16b..bf7a4c0551abe35e3572459576181341ff33c8a7 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/dpotrf.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/dpotrf.sirocco @@ -7,12 +7,12 @@ # number of combinations 5 #################### -# COMB_1 +# COMB_0 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cuda0_impl0 (Comb1) +# Model for cuda0_impl0 (Comb0) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,19 +36,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -d48836c9 16588800 0.000000e+00 1.927652e+04 3.502107e+03 9.059964e+05 1.804090e+10 47 -901ddf46 7372800 0.000000e+00 1.229853e+04 1.768810e+03 6.149267e+05 7.719131e+09 50 -5c532f3b 51200 0.000000e+00 2.101636e+03 2.756395e+02 1.303014e+05 2.785568e+08 62 -69439308 1548800 0.000000e+00 4.903930e+03 3.193701e+02 2.550044e+05 1.255828e+09 52 a0677317 29491200 0.000000e+00 2.463348e+04 2.985247e+03 1.576542e+06 3.940607e+10 64 +69439308 1548800 0.000000e+00 4.903930e+03 3.193701e+02 2.550044e+05 1.255828e+09 52 +5c532f3b 51200 0.000000e+00 2.101636e+03 2.756395e+02 1.303014e+05 2.785568e+08 62 +901ddf46 7372800 0.000000e+00 1.229853e+04 1.768810e+03 6.149267e+05 7.719131e+09 50 +d48836c9 16588800 0.000000e+00 1.927652e+04 3.502107e+03 9.059964e+05 1.804090e+10 47 #################### -# COMB_2 +# COMB_3 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -62,7 +62,7 @@ a0677317 29491200 0.000000e+00 2.463348e+04 2.985247e+03 1.57654 # number of implementations 1 ##### -# Model for cuda2_impl0 (Comb2) +# Model for cuda2_impl0 (Comb3) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -72,19 +72,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -d48836c9 16588800 0.000000e+00 1.940135e+04 4.068762e+03 6.984485e+05 1.414682e+10 36 -901ddf46 7372800 0.000000e+00 1.194505e+04 1.978674e+03 3.822416e+05 4.691180e+09 32 -5c532f3b 51200 0.000000e+00 2.114340e+03 2.957222e+02 1.268604e+05 2.734732e+08 60 -69439308 1548800 0.000000e+00 5.046035e+03 5.465455e+02 1.766112e+05 9.016412e+08 35 a0677317 29491200 0.000000e+00 2.329330e+04 2.356823e+03 1.094785e+06 2.576222e+10 47 +69439308 1548800 0.000000e+00 5.046035e+03 5.465455e+02 1.766112e+05 9.016412e+08 35 +5c532f3b 51200 0.000000e+00 2.114340e+03 2.957222e+02 1.268604e+05 2.734732e+08 60 +901ddf46 7372800 0.000000e+00 1.194505e+04 1.978674e+03 3.822416e+05 4.691180e+09 32 +d48836c9 16588800 0.000000e+00 1.940135e+04 4.068762e+03 6.984485e+05 1.414682e+10 36 #################### -# COMB_3 +# COMB_4 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -98,7 +98,7 @@ a0677317 29491200 0.000000e+00 2.329330e+04 2.356823e+03 1.09478 # number of implementations 1 ##### -# Model for cuda3_impl0 (Comb3) +# Model for cuda3_impl0 (Comb4) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -108,19 +108,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -d48836c9 16588800 0.000000e+00 1.848606e+04 2.220768e+03 5.545817e+05 1.039998e+10 30 -901ddf46 7372800 0.000000e+00 1.174012e+04 1.793250e+03 3.991641e+05 4.795570e+09 34 -5c532f3b 51200 0.000000e+00 2.130842e+03 3.337398e+02 1.576823e+05 3.442383e+08 74 -69439308 1548800 0.000000e+00 5.078755e+03 5.140761e+02 2.437802e+05 1.250785e+09 48 a0677317 29491200 0.000000e+00 2.451908e+04 3.131104e+03 5.884580e+05 1.466374e+10 24 +69439308 1548800 0.000000e+00 5.078755e+03 5.140761e+02 2.437802e+05 1.250785e+09 48 +5c532f3b 51200 0.000000e+00 2.130842e+03 3.337398e+02 1.576823e+05 3.442383e+08 74 +901ddf46 7372800 0.000000e+00 1.174012e+04 1.793250e+03 3.991641e+05 4.795570e+09 34 +d48836c9 16588800 0.000000e+00 1.848606e+04 2.220768e+03 5.545817e+05 1.039998e+10 30 #################### -# COMB_0 +# COMB_1 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -134,7 +134,7 @@ a0677317 29491200 0.000000e+00 2.451908e+04 3.131104e+03 5.88458 # number of implementations 1 ##### -# Model for cuda1_impl0 (Comb0) +# Model for cuda1_impl0 (Comb1) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -144,19 +144,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -d48836c9 16588800 0.000000e+00 1.807832e+04 2.125912e+03 7.412111e+05 1.358515e+10 41 -5c532f3b 51200 0.000000e+00 2.127387e+03 2.904710e+02 1.404075e+05 3.042697e+08 66 -901ddf46 7372800 0.000000e+00 1.177248e+04 1.259016e+03 4.002642e+05 4.765995e+09 34 -69439308 1548800 0.000000e+00 4.993474e+03 5.010162e+02 2.546672e+05 1.284476e+09 51 a0677317 29491200 0.000000e+00 2.412058e+04 2.891586e+03 1.206029e+06 2.950819e+10 50 +69439308 1548800 0.000000e+00 4.993474e+03 5.010162e+02 2.546672e+05 1.284476e+09 51 +901ddf46 7372800 0.000000e+00 1.177248e+04 1.259016e+03 4.002642e+05 4.765995e+09 34 +5c532f3b 51200 0.000000e+00 2.127387e+03 2.904710e+02 1.404075e+05 3.042697e+08 66 +d48836c9 16588800 0.000000e+00 1.807832e+04 2.125912e+03 7.412111e+05 1.358515e+10 41 #################### -# COMB_4 +# COMB_2 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 0 #################### # DEV_0 @@ -170,7 +170,7 @@ a0677317 29491200 0.000000e+00 2.412058e+04 2.891586e+03 1.20602 # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb4) +# Model for cpu0_impl0 (Comb2) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -180,9 +180,9 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -d48836c9 16588800 0.000000e+00 3.521786e+04 3.719259e+03 2.183508e+06 7.775611e+10 62 -69439308 1548800 0.000000e+00 1.965233e+03 3.623952e+02 7.271360e+04 1.477584e+08 37 -901ddf46 7372800 0.000000e+00 1.125962e+04 1.634952e+03 5.855000e+05 6.731505e+09 52 -5c532f3b 51200 0.000000e+00 4.077398e+01 4.776338e+00 2.487213e+03 1.028052e+05 61 a0677317 29491200 0.000000e+00 7.956441e+04 8.548094e+03 5.410380e+06 4.354424e+11 68 +5c532f3b 51200 0.000000e+00 4.077398e+01 4.776338e+00 2.487213e+03 1.028052e+05 61 +901ddf46 7372800 0.000000e+00 1.114621e+04 1.459505e+03 3.042915e+06 3.449850e+10 273 +69439308 1548800 0.000000e+00 1.965233e+03 3.623952e+02 7.271360e+04 1.477584e+08 37 +d48836c9 16588800 0.000000e+00 3.521786e+04 3.719259e+03 2.183508e+06 7.775611e+10 62 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/dsyrk.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/dsyrk.sirocco index 66a3a6695d396095bd3248adadafa9c1f5ed52be..96de8105d90a0747521872794e2445e3e1134662 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/dsyrk.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/dsyrk.sirocco @@ -7,12 +7,12 @@ # number of combinations 5 #################### -# COMB_4 +# COMB_2 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 0 #################### # DEV_0 @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb4) +# Model for cpu0_impl0 (Comb2) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,19 +36,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 3.671926e+04 6.264954e+03 4.868973e+07 1.839896e+12 1326 -6ab5620 102400 0.000000e+00 3.685897e+01 6.246340e+00 2.291522e+05 8.688881e+06 6217 -34b59a6f 3097600 0.000000e+00 3.808134e+03 7.661841e+02 8.320772e+06 3.296929e+10 2185 -50116ceb 33177600 0.000000e+00 1.197936e+05 2.202875e+04 9.260042e+07 1.146804e+13 773 ae9b1fd5 58982400 0.000000e+00 2.500810e+05 3.400228e+04 1.668040e+08 4.248568e+13 667 +50116ceb 33177600 0.000000e+00 1.197936e+05 2.202875e+04 9.260042e+07 1.146804e+13 773 +34b59a6f 3097600 0.000000e+00 3.808134e+03 7.661841e+02 8.320772e+06 3.296929e+10 2185 +06ab5620 102400 0.000000e+00 3.685897e+01 6.246340e+00 2.291522e+05 8.688881e+06 6217 +a1187f60 14745600 0.000000e+00 3.750924e+04 7.601332e+03 1.832702e+08 7.156640e+12 4886 #################### -# COMB_2 +# COMB_0 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -62,7 +62,7 @@ ae9b1fd5 58982400 0.000000e+00 2.500810e+05 3.400228e+04 1.66804 # number of implementations 1 ##### -# Model for cuda0_impl0 (Comb2) +# Model for cuda0_impl0 (Comb0) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -72,11 +72,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 1.225366e+03 1.055960e+02 2.940879e+05 3.630415e+08 240 -6ab5620 102400 0.000000e+00 6.983900e+01 1.260925e+01 8.380680e+02 6.043775e+04 12 -34b59a6f 3097600 0.000000e+00 2.036579e+02 2.324388e+01 5.885713e+04 1.214286e+07 289 -50116ceb 33177600 0.000000e+00 3.666675e+03 2.800845e+02 7.150015e+05 2.636975e+09 195 ae9b1fd5 58982400 0.000000e+00 7.329175e+03 4.915258e+02 1.693039e+06 1.246439e+10 231 +50116ceb 33177600 0.000000e+00 3.666675e+03 2.800845e+02 7.150015e+05 2.636975e+09 195 +34b59a6f 3097600 0.000000e+00 2.036579e+02 2.324388e+01 5.885713e+04 1.214286e+07 289 +06ab5620 102400 0.000000e+00 6.983900e+01 1.260925e+01 8.380680e+02 6.043775e+04 12 +a1187f60 14745600 0.000000e+00 1.243616e+03 7.741866e+01 8.033757e+05 1.002962e+09 646 #################### # COMB_1 @@ -84,7 +84,7 @@ ae9b1fd5 58982400 0.000000e+00 7.329175e+03 4.915258e+02 1.69303 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -108,19 +108,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 1.251704e+03 9.523925e+01 2.928988e+05 3.687452e+08 234 -6ab5620 102400 0.000000e+00 7.712755e+01 1.662959e+01 8.484030e+02 6.847722e+04 11 -34b59a6f 3097600 0.000000e+00 2.078053e+02 2.202377e+01 5.402939e+04 1.135371e+07 260 -50116ceb 33177600 0.000000e+00 3.720059e+03 2.853133e+02 8.295732e+05 3.104214e+09 223 ae9b1fd5 58982400 0.000000e+00 7.383529e+03 3.772204e+02 1.580075e+06 1.169698e+10 214 +50116ceb 33177600 0.000000e+00 3.720059e+03 2.853133e+02 8.295732e+05 3.104214e+09 223 +34b59a6f 3097600 0.000000e+00 2.078053e+02 2.202377e+01 5.402939e+04 1.135371e+07 260 +06ab5620 102400 0.000000e+00 7.712755e+01 1.662959e+01 8.484030e+02 6.847722e+04 11 +a1187f60 14745600 0.000000e+00 1.244957e+03 7.232457e+01 8.042425e+05 1.004627e+09 646 #################### -# COMB_0 +# COMB_3 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -134,7 +134,7 @@ ae9b1fd5 58982400 0.000000e+00 7.383529e+03 3.772204e+02 1.58007 # number of implementations 1 ##### -# Model for cuda2_impl0 (Comb0) +# Model for cuda2_impl0 (Comb3) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -144,19 +144,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 1.252296e+03 1.169465e+02 2.955417e+05 3.733333e+08 236 -6ab5620 102400 0.000000e+00 7.692512e+01 1.259036e+01 1.230802e+03 9.721587e+04 16 -34b59a6f 3097600 0.000000e+00 2.079242e+02 2.471557e+01 6.029802e+04 1.271457e+07 290 -50116ceb 33177600 0.000000e+00 3.744678e+03 3.676926e+02 7.938716e+05 3.001455e+09 212 ae9b1fd5 58982400 0.000000e+00 7.454402e+03 4.686588e+02 1.669786e+06 1.249646e+10 224 +50116ceb 33177600 0.000000e+00 3.744678e+03 3.676926e+02 7.938716e+05 3.001455e+09 212 +34b59a6f 3097600 0.000000e+00 2.079242e+02 2.471557e+01 6.029802e+04 1.271457e+07 290 +06ab5620 102400 0.000000e+00 7.692512e+01 1.259036e+01 1.230802e+03 9.721587e+04 16 +a1187f60 14745600 0.000000e+00 1.252296e+03 1.169465e+02 2.955417e+05 3.733333e+08 236 #################### -# COMB_3 +# COMB_4 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -170,7 +170,7 @@ ae9b1fd5 58982400 0.000000e+00 7.454402e+03 4.686588e+02 1.66978 # number of implementations 1 ##### -# Model for cuda3_impl0 (Comb3) +# Model for cuda3_impl0 (Comb4) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -180,9 +180,9 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 1.227027e+03 1.116503e+02 2.908055e+05 3.597807e+08 237 -6ab5620 102400 0.000000e+00 7.100592e+01 1.678181e+01 8.520710e+02 6.388163e+04 12 -34b59a6f 3097600 0.000000e+00 2.027211e+02 2.465123e+01 5.331564e+04 1.096802e+07 263 -50116ceb 33177600 0.000000e+00 3.582949e+03 2.599939e+02 6.341819e+05 2.284206e+09 177 ae9b1fd5 58982400 0.000000e+00 7.239223e+03 4.831241e+02 1.360974e+06 9.896276e+09 188 +50116ceb 33177600 0.000000e+00 3.582949e+03 2.599939e+02 6.341819e+05 2.284206e+09 177 +34b59a6f 3097600 0.000000e+00 2.027211e+02 2.465123e+01 5.331564e+04 1.096802e+07 263 +06ab5620 102400 0.000000e+00 7.100592e+01 1.678181e+01 8.520710e+02 6.388163e+04 12 +a1187f60 14745600 0.000000e+00 1.227027e+03 1.116503e+02 2.908055e+05 3.597807e+08 237 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/dtrmm.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/dtrmm.sirocco new file mode 100644 index 0000000000000000000000000000000000000000..ebe055f41eb63d80c63c6abac37bd0e498f7fb51 --- /dev/null +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/dtrmm.sirocco @@ -0,0 +1,104 @@ +################## +# Performance Model Version +45 + +#################### +# COMBs +# number of combinations +3 +#################### +# COMB_2 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +0 +#################### +# DEV_0 +# device id +0 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cpu0_impl0 (Comb2) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +a1187f60 14745600 0.000000e+00 3.793090e+04 6.009200e+03 7.665835e+07 2.980700e+12 2021 + +#################### +# COMB_1 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +1 +#################### +# DEV_0 +# device id +1 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cuda1_impl0 (Comb1) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +a1187f60 14745600 0.000000e+00 1.955835e+03 7.118363e+01 5.613246e+05 1.099313e+09 287 + +#################### +# COMB_0 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +1 +#################### +# DEV_0 +# device id +0 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cuda0_impl0 (Comb0) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +a1187f60 14745600 0.000000e+00 1.964850e+03 6.368649e+01 5.069313e+05 9.970904e+08 258 + diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/dtrsm.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/dtrsm.sirocco index 7a19b71921b93c897aea1065a2f695ef56eda913..6c83e4ed2cb18d23133a25cdcaddc195e34cb90a 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/dtrsm.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/dtrsm.sirocco @@ -7,12 +7,12 @@ # number of combinations 5 #################### -# COMB_4 +# COMB_2 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 0 #################### # DEV_0 @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb4) +# Model for cpu0_impl0 (Comb2) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,19 +36,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 3.428845e+04 6.014106e+03 4.670087e+07 1.650564e+12 1362 -6ab5620 102400 0.000000e+00 4.315566e+01 9.048618e+00 2.734774e+05 1.232096e+07 6337 -34b59a6f 3097600 0.000000e+00 3.493952e+03 6.537880e+02 8.623074e+06 3.118353e+10 2468 -50116ceb 33177600 0.000000e+00 1.022247e+05 1.393583e+04 9.251336e+07 9.632909e+12 905 ae9b1fd5 58982400 0.000000e+00 2.235792e+05 2.485110e+04 1.562819e+08 3.537307e+13 699 +50116ceb 33177600 0.000000e+00 1.022247e+05 1.393583e+04 9.251336e+07 9.632909e+12 905 +34b59a6f 3097600 0.000000e+00 3.493952e+03 6.537880e+02 8.623074e+06 3.118353e+10 2468 +06ab5620 102400 0.000000e+00 4.315566e+01 9.048618e+00 2.734774e+05 1.232096e+07 6337 +a1187f60 14745600 0.000000e+00 3.930876e+04 9.409133e+03 1.559379e+08 6.480929e+12 3967 #################### -# COMB_0 +# COMB_3 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -62,7 +62,7 @@ ae9b1fd5 58982400 0.000000e+00 2.235792e+05 2.485110e+04 1.56281 # number of implementations 1 ##### -# Model for cuda2_impl0 (Comb0) +# Model for cuda2_impl0 (Comb3) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -72,11 +72,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 3.276119e+03 3.115623e+02 8.157537e+05 2.696677e+09 249 -6ab5620 102400 0.000000e+00 5.400294e+01 1.227261e+01 9.720530e+02 5.520483e+04 18 -34b59a6f 3097600 0.000000e+00 1.059018e+03 1.208041e+02 2.509872e+05 2.692586e+08 237 -50116ceb 33177600 0.000000e+00 7.073186e+03 6.003818e+02 1.082197e+06 7.709734e+09 153 ae9b1fd5 58982400 0.000000e+00 1.275670e+04 1.000532e+03 2.462043e+06 3.160076e+10 193 +50116ceb 33177600 0.000000e+00 7.073186e+03 6.003818e+02 1.082197e+06 7.709734e+09 153 +34b59a6f 3097600 0.000000e+00 1.059018e+03 1.208041e+02 2.509872e+05 2.692586e+08 237 +06ab5620 102400 0.000000e+00 5.400294e+01 1.227261e+01 9.720530e+02 5.520483e+04 18 +a1187f60 14745600 0.000000e+00 3.276119e+03 3.115623e+02 8.157537e+05 2.696677e+09 249 #################### # COMB_1 @@ -84,7 +84,7 @@ ae9b1fd5 58982400 0.000000e+00 1.275670e+04 1.000532e+03 2.46204 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -108,19 +108,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 3.300871e+03 2.581032e+02 6.568733e+05 2.181511e+09 199 -6ab5620 102400 0.000000e+00 1.639117e+02 3.695355e+01 1.966940e+03 3.387912e+05 12 -34b59a6f 3097600 0.000000e+00 1.035320e+03 1.634986e+02 2.370884e+05 2.515840e+08 229 -50116ceb 33177600 0.000000e+00 7.177520e+03 7.682785e+02 1.342196e+06 9.744017e+09 187 ae9b1fd5 58982400 0.000000e+00 1.284895e+04 1.132613e+03 2.479846e+06 3.211099e+10 193 +50116ceb 33177600 0.000000e+00 7.177520e+03 7.682785e+02 1.342196e+06 9.744017e+09 187 +34b59a6f 3097600 0.000000e+00 1.035320e+03 1.634986e+02 2.370884e+05 2.515840e+08 229 +06ab5620 102400 0.000000e+00 1.639117e+02 3.695355e+01 1.966940e+03 3.387912e+05 12 +a1187f60 14745600 0.000000e+00 3.180532e+03 1.979664e+02 5.340114e+06 1.705021e+10 1679 #################### -# COMB_3 +# COMB_4 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -134,7 +134,7 @@ ae9b1fd5 58982400 0.000000e+00 1.284895e+04 1.132613e+03 2.47984 # number of implementations 1 ##### -# Model for cuda3_impl0 (Comb3) +# Model for cuda3_impl0 (Comb4) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -144,19 +144,19 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 3.224312e+03 2.905618e+02 8.576671e+05 2.787844e+09 266 -6ab5620 102400 0.000000e+00 1.238758e+02 2.453256e+01 1.238758e+03 1.594706e+05 10 -34b59a6f 3097600 0.000000e+00 1.043152e+03 1.402538e+02 2.597449e+05 2.758515e+08 249 -50116ceb 33177600 0.000000e+00 6.893239e+03 4.958485e+02 1.557872e+06 1.079435e+10 226 ae9b1fd5 58982400 0.000000e+00 1.233960e+04 7.352896e+02 2.924485e+06 3.621511e+10 237 +50116ceb 33177600 0.000000e+00 6.893239e+03 4.958485e+02 1.557872e+06 1.079435e+10 226 +34b59a6f 3097600 0.000000e+00 1.043152e+03 1.402538e+02 2.597449e+05 2.758515e+08 249 +06ab5620 102400 0.000000e+00 1.238758e+02 2.453256e+01 1.238758e+03 1.594706e+05 10 +a1187f60 14745600 0.000000e+00 3.224312e+03 2.905618e+02 8.576671e+05 2.787844e+09 266 #################### -# COMB_2 +# COMB_0 # number of types devices 1 #################### # DEV_0 -# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, SCC - 4) +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) 1 #################### # DEV_0 @@ -170,7 +170,7 @@ ae9b1fd5 58982400 0.000000e+00 1.233960e+04 7.352896e+02 2.92448 # number of implementations 1 ##### -# Model for cuda0_impl0 (Comb2) +# Model for cuda0_impl0 (Comb0) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -180,9 +180,9 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a1187f60 14745600 0.000000e+00 3.226919e+03 2.819444e+02 8.357719e+05 2.717557e+09 259 -6ab5620 102400 0.000000e+00 7.576700e+01 1.461196e+01 1.288039e+03 1.012205e+05 17 -34b59a6f 3097600 0.000000e+00 1.050995e+03 1.429382e+02 2.385759e+05 2.553801e+08 227 -50116ceb 33177600 0.000000e+00 7.077068e+03 7.343420e+02 1.288026e+06 9.213596e+09 182 ae9b1fd5 58982400 0.000000e+00 1.259917e+04 1.081288e+03 2.797016e+06 3.549963e+10 222 +50116ceb 33177600 0.000000e+00 7.077068e+03 7.343420e+02 1.288026e+06 9.213596e+09 182 +34b59a6f 3097600 0.000000e+00 1.050995e+03 1.429382e+02 2.385759e+05 2.553801e+08 227 +06ab5620 102400 0.000000e+00 7.576700e+01 1.461196e+01 1.288039e+03 1.012205e+05 17 +a1187f60 14745600 0.000000e+00 3.191368e+03 1.968492e+02 5.205120e+06 1.667465e+10 1631 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/dtrtri.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/dtrtri.sirocco new file mode 100644 index 0000000000000000000000000000000000000000..841ec75946f4de305cab428dfeb0a7ce8e8866b2 --- /dev/null +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/dtrtri.sirocco @@ -0,0 +1,40 @@ +################## +# Performance Model Version +45 + +#################### +# COMBs +# number of combinations +1 +#################### +# COMB_2 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +0 +#################### +# DEV_0 +# device id +0 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cpu0_impl0 (Comb2) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +901ddf46 7372800 0.000000e+00 2.555721e+04 5.276570e+03 3.578009e+06 9.534185e+10 140 + diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/sgemm.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/sgemm.sirocco index b093180c891e0e63b83da2d89c33ca20cbfb96e5..94e8a05dce7f846abcdfae76cdc94ddb04bac27b 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/sgemm.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/sgemm.sirocco @@ -7,7 +7,7 @@ # number of combinations 5 #################### -# COMB_0 +# COMB_2 # number of types devices 1 #################### @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb0) +# Model for cpu0_impl0 (Comb2) # number of entries 6 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,12 +36,12 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -f74a4f19 44236800 0.000000e+00 2.633445e+05 5.249424e+04 6.322902e+08 1.731265e+14 2401 -dc7b85e2 24883200 0.000000e+00 1.120333e+05 2.070542e+04 2.092782e+08 2.424696e+13 1868 -744ff412 11059200 0.000000e+00 3.623402e+04 6.546678e+03 1.580528e+08 5.913839e+12 4362 -9427311d 2323200 0.000000e+00 3.732336e+03 6.841164e+02 3.950678e+07 1.524065e+11 10585 -b63efa5a 76800 0.000000e+00 3.040738e+01 3.904133e+00 2.440131e+06 7.542116e+07 80248 84e38ff9 1228800 0.000000e+00 1.962744e+03 4.745716e+02 4.710585e+04 9.786197e+07 24 +b63efa5a 76800 0.000000e+00 3.040738e+01 3.904133e+00 2.440131e+06 7.542116e+07 80248 +9427311d 2323200 0.000000e+00 3.732336e+03 6.841164e+02 3.950678e+07 1.524065e+11 10585 +744ff412 11059200 0.000000e+00 3.776331e+04 7.086585e+03 2.317157e+08 9.058498e+12 6136 +dc7b85e2 24883200 0.000000e+00 1.120333e+05 2.070542e+04 2.092782e+08 2.424696e+13 1868 +f74a4f19 44236800 0.000000e+00 2.633445e+05 5.249424e+04 6.322902e+08 1.731265e+14 2401 #################### # COMB_3 @@ -73,14 +73,14 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -f74a4f19 44236800 0.000000e+00 6.265723e+03 4.806927e+02 8.577149e+07 5.405835e+11 13689 -dc7b85e2 24883200 0.000000e+00 2.982067e+03 1.577603e+02 2.606923e+07 7.795776e+10 8742 -744ff412 11059200 0.000000e+00 1.027724e+03 6.416246e+01 1.796770e+07 1.853781e+10 17483 -9427311d 2323200 0.000000e+00 1.681587e+02 2.185477e+01 2.332192e+06 3.988026e+08 13869 b63efa5a 76800 0.000000e+00 4.561876e+01 1.081980e+01 6.934052e+03 3.341172e+05 152 +9427311d 2323200 0.000000e+00 1.681587e+02 2.185477e+01 2.332192e+06 3.988026e+08 13869 +744ff412 11059200 0.000000e+00 1.027724e+03 6.416246e+01 1.796770e+07 1.853781e+10 17483 +dc7b85e2 24883200 0.000000e+00 2.982067e+03 1.577603e+02 2.606923e+07 7.795776e+10 8742 +f74a4f19 44236800 0.000000e+00 6.265723e+03 4.806927e+02 8.577149e+07 5.405835e+11 13689 #################### -# COMB_1 +# COMB_0 # number of types devices 1 #################### @@ -99,7 +99,7 @@ b63efa5a 76800 0.000000e+00 4.561876e+01 1.081980e+01 6.93405 # number of implementations 1 ##### -# Model for cuda0_impl0 (Comb1) +# Model for cuda0_impl0 (Comb0) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -109,14 +109,14 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -f74a4f19 44236800 0.000000e+00 6.411657e+03 4.740807e+02 8.565973e+07 5.522235e+11 13360 -dc7b85e2 24883200 0.000000e+00 2.990259e+03 1.556149e+02 2.622756e+07 7.863960e+10 8771 -744ff412 11059200 0.000000e+00 1.026941e+03 6.843339e+01 1.818713e+07 1.876005e+10 17710 -9427311d 2323200 0.000000e+00 1.678676e+02 2.188969e+01 2.444656e+06 4.173565e+08 14563 b63efa5a 76800 0.000000e+00 5.712863e+01 1.016253e+01 2.182314e+04 1.286178e+06 382 +9427311d 2323200 0.000000e+00 1.678676e+02 2.188969e+01 2.444656e+06 4.173565e+08 14563 +744ff412 11059200 0.000000e+00 9.242827e+02 1.465376e+02 4.502181e+07 4.265884e+10 48710 +dc7b85e2 24883200 0.000000e+00 2.990259e+03 1.556149e+02 2.622756e+07 7.863960e+10 8771 +f74a4f19 44236800 0.000000e+00 6.411657e+03 4.740807e+02 8.565973e+07 5.522235e+11 13360 #################### -# COMB_2 +# COMB_1 # number of types devices 1 #################### @@ -135,7 +135,7 @@ b63efa5a 76800 0.000000e+00 5.712863e+01 1.016253e+01 2.18231 # number of implementations 1 ##### -# Model for cuda1_impl0 (Comb2) +# Model for cuda1_impl0 (Comb1) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -145,11 +145,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -f74a4f19 44236800 0.000000e+00 6.216224e+03 4.119793e+02 8.361443e+07 5.220491e+11 13451 -dc7b85e2 24883200 0.000000e+00 3.034574e+03 1.617988e+02 2.660715e+07 8.097090e+10 8768 -744ff412 11059200 0.000000e+00 1.042160e+03 6.826839e+01 1.799915e+07 1.883849e+10 17271 -9427311d 2323200 0.000000e+00 1.706067e+02 2.205696e+01 2.522932e+06 4.376237e+08 14788 b63efa5a 76800 0.000000e+00 4.872868e+01 7.797709e+00 3.376898e+04 1.687655e+06 693 +9427311d 2323200 0.000000e+00 1.706067e+02 2.205696e+01 2.522932e+06 4.376237e+08 14788 +744ff412 11059200 0.000000e+00 9.233103e+02 1.525748e+02 4.493289e+07 4.261988e+10 48665 +dc7b85e2 24883200 0.000000e+00 3.034574e+03 1.617988e+02 2.660715e+07 8.097090e+10 8768 +f74a4f19 44236800 0.000000e+00 6.216224e+03 4.119793e+02 8.361443e+07 5.220491e+11 13451 #################### # COMB_4 @@ -181,9 +181,9 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -f74a4f19 44236800 0.000000e+00 6.193669e+03 4.216424e+02 8.780145e+07 5.463333e+11 14176 -dc7b85e2 24883200 0.000000e+00 3.030444e+03 1.493135e+02 2.715884e+07 8.250317e+10 8962 -744ff412 11059200 0.000000e+00 1.042297e+03 6.563263e+01 1.817663e+07 1.902057e+10 17439 -9427311d 2323200 0.000000e+00 1.710487e+02 2.221991e+01 2.555297e+06 4.444559e+08 14939 b63efa5a 76800 0.000000e+00 5.246265e+01 9.697468e+00 1.453215e+04 7.884446e+05 277 +9427311d 2323200 0.000000e+00 1.710487e+02 2.221991e+01 2.555297e+06 4.444559e+08 14939 +744ff412 11059200 0.000000e+00 1.042297e+03 6.563263e+01 1.817663e+07 1.902057e+10 17439 +dc7b85e2 24883200 0.000000e+00 3.030444e+03 1.493135e+02 2.715884e+07 8.250317e+10 8962 +f74a4f19 44236800 0.000000e+00 6.193669e+03 4.216424e+02 8.780145e+07 5.463333e+11 14176 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/slauum.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/slauum.sirocco new file mode 100644 index 0000000000000000000000000000000000000000..7bcb5922babe8fca6e3a6ba283626916fc71fd0d --- /dev/null +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/slauum.sirocco @@ -0,0 +1,40 @@ +################## +# Performance Model Version +45 + +#################### +# COMBs +# number of combinations +1 +#################### +# COMB_2 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +0 +#################### +# DEV_0 +# device id +0 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cpu0_impl0 (Comb2) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +901ddf46 3686400 0.000000e+00 7.577730e+03 1.166374e+03 1.932321e+06 1.498952e+10 255 + diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/splgsy.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/splgsy.sirocco index c4530465ccc3e2a790829d43f79222c8e9336d37..5d67ef47fb4fc040e061c3f7a3702879ba5d01f6 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/splgsy.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/splgsy.sirocco @@ -7,7 +7,7 @@ # number of combinations 1 #################### -# COMB_0 +# COMB_2 # number of types devices 1 #################### @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb0) +# Model for cpu0_impl0 (Comb2) # number of entries 8 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,12 +36,12 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a0677317 14745600 0.000000e+00 3.545004e+04 1.130964e+03 1.871762e+07 6.642158e+11 528 -d48836c9 8294400 0.000000e+00 1.998968e+04 3.810275e+02 4.657595e+06 9.313763e+10 233 -901ddf46 3686400 0.000000e+00 8.939938e+03 2.153220e+02 1.231923e+07 1.101971e+11 1378 -69439308 774400 0.000000e+00 2.302314e+03 5.466484e+02 1.138264e+07 2.768379e+10 4944 -5c532f3b 25600 0.000000e+00 8.731844e+01 8.227614e+00 1.389149e+06 1.223753e+08 15909 -3442589a 6400 0.000000e+00 3.342405e+01 1.257385e+01 6.684810e+02 2.550538e+04 20 -30b2a96f 51200 0.000000e+00 1.475864e+02 1.554810e+01 1.136415e+04 1.695808e+06 77 3fc0fdcd 409600 0.000000e+00 8.902086e+02 1.436824e+02 1.370921e+05 1.252199e+08 154 +30b2a96f 51200 0.000000e+00 1.475864e+02 1.554810e+01 1.136415e+04 1.695808e+06 77 +3442589a 6400 0.000000e+00 3.342405e+01 1.257385e+01 6.684810e+02 2.550538e+04 20 +5c532f3b 25600 0.000000e+00 8.731844e+01 8.227614e+00 1.389149e+06 1.223753e+08 15909 +69439308 774400 0.000000e+00 2.302314e+03 5.466484e+02 1.138264e+07 2.768379e+10 4944 +901ddf46 3686400 0.000000e+00 8.479906e+03 1.531934e+03 3.893125e+07 3.409076e+11 4591 +d48836c9 8294400 0.000000e+00 1.998968e+04 3.810275e+02 4.657595e+06 9.313763e+10 233 +a0677317 14745600 0.000000e+00 3.545004e+04 1.130964e+03 1.871762e+07 6.642158e+11 528 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/spotrf.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/spotrf.sirocco index d0b792059e70b4fd554ef3d3cda822f17510b3f8..8829d6ea39b999803a57effcb4b21886529bff59 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/spotrf.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/spotrf.sirocco @@ -7,7 +7,7 @@ # number of combinations 5 #################### -# COMB_1 +# COMB_0 # number of types devices 1 #################### @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cuda0_impl0 (Comb1) +# Model for cuda0_impl0 (Comb0) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,14 +36,14 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a0677317 14745600 0.000000e+00 1.725922e+04 2.264939e+03 1.052812e+06 1.848364e+10 61 -d48836c9 8294400 0.000000e+00 1.618055e+04 2.196987e+03 7.766664e+05 1.279857e+10 48 -69439308 774400 0.000000e+00 3.928085e+03 4.967651e+02 1.414111e+05 5.643587e+08 36 -5c532f3b 25600 0.000000e+00 2.070017e+03 2.556262e+02 1.283410e+05 2.697195e+08 62 901ddf46 3686400 0.000000e+00 9.570246e+03 1.453408e+03 5.263635e+05 5.153611e+09 55 +5c532f3b 25600 0.000000e+00 2.070017e+03 2.556262e+02 1.283410e+05 2.697195e+08 62 +69439308 774400 0.000000e+00 3.928085e+03 4.967651e+02 1.414111e+05 5.643587e+08 36 +d48836c9 8294400 0.000000e+00 1.618055e+04 2.196987e+03 7.766664e+05 1.279857e+10 48 +a0677317 14745600 0.000000e+00 1.725922e+04 2.264939e+03 1.052812e+06 1.848364e+10 61 #################### -# COMB_2 +# COMB_1 # number of types devices 1 #################### @@ -62,7 +62,7 @@ d48836c9 8294400 0.000000e+00 1.618055e+04 2.196987e+03 7.76666 # number of implementations 1 ##### -# Model for cuda1_impl0 (Comb2) +# Model for cuda1_impl0 (Comb1) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -72,11 +72,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a0677317 14745600 0.000000e+00 1.689494e+04 1.439976e+03 1.081276e+06 1.840081e+10 64 -d48836c9 8294400 0.000000e+00 1.549399e+04 1.742327e+03 8.211813e+05 1.288426e+10 53 -69439308 774400 0.000000e+00 3.925510e+03 5.372277e+02 1.334673e+05 5.337401e+08 34 -5c532f3b 25600 0.000000e+00 2.102179e+03 3.123586e+02 1.303351e+05 2.800370e+08 62 901ddf46 3686400 0.000000e+00 9.539420e+03 1.257811e+03 4.292739e+05 4.166218e+09 45 +5c532f3b 25600 0.000000e+00 2.102179e+03 3.123586e+02 1.303351e+05 2.800370e+08 62 +69439308 774400 0.000000e+00 3.925510e+03 5.372277e+02 1.334673e+05 5.337401e+08 34 +d48836c9 8294400 0.000000e+00 1.549399e+04 1.742327e+03 8.211813e+05 1.288426e+10 53 +a0677317 14745600 0.000000e+00 1.689494e+04 1.439976e+03 1.081276e+06 1.840081e+10 64 #################### # COMB_3 @@ -108,11 +108,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a0677317 14745600 0.000000e+00 1.679050e+04 1.864821e+03 5.037149e+05 8.561949e+09 30 -d48836c9 8294400 0.000000e+00 1.650986e+04 2.615948e+03 4.457661e+05 7.544300e+09 27 -69439308 774400 0.000000e+00 4.072426e+03 6.492225e+02 1.384625e+05 5.782087e+08 34 +901ddf46 3686400 0.000000e+00 9.596093e+03 1.252122e+03 2.686906e+05 2.622279e+09 28 5c532f3b 25600 0.000000e+00 2.106632e+03 2.976700e+02 1.390377e+05 2.987493e+08 66 -901ddf46 3686400 0.000000e+00 9.596093e+03 1.252122e+03 2.686906e+05 2.622279e+09 28 +69439308 774400 0.000000e+00 4.072426e+03 6.492225e+02 1.384625e+05 5.782087e+08 34 +d48836c9 8294400 0.000000e+00 1.650986e+04 2.615948e+03 4.457661e+05 7.544300e+09 27 +a0677317 14745600 0.000000e+00 1.679050e+04 1.864821e+03 5.037149e+05 8.561949e+09 30 #################### # COMB_4 @@ -144,14 +144,14 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a0677317 14745600 0.000000e+00 1.640619e+04 1.490045e+03 6.398413e+05 1.058394e+10 39 -d48836c9 8294400 0.000000e+00 1.538862e+04 2.385658e+03 6.155449e+05 9.700043e+09 40 -69439308 774400 0.000000e+00 4.069845e+03 5.836273e+02 1.627938e+05 6.761704e+08 40 -5c532f3b 25600 0.000000e+00 2.134159e+03 2.682985e+02 1.323178e+05 2.868503e+08 62 901ddf46 3686400 0.000000e+00 9.181625e+03 8.594056e+02 3.580834e+05 3.316592e+09 39 +5c532f3b 25600 0.000000e+00 2.134159e+03 2.682985e+02 1.323178e+05 2.868503e+08 62 +69439308 774400 0.000000e+00 4.069845e+03 5.836273e+02 1.627938e+05 6.761704e+08 40 +d48836c9 8294400 0.000000e+00 1.538862e+04 2.385658e+03 6.155449e+05 9.700043e+09 40 +a0677317 14745600 0.000000e+00 1.640619e+04 1.490045e+03 6.398413e+05 1.058394e+10 39 #################### -# COMB_0 +# COMB_2 # number of types devices 1 #################### @@ -170,7 +170,7 @@ d48836c9 8294400 0.000000e+00 1.538862e+04 2.385658e+03 6.15544 # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb0) +# Model for cpu0_impl0 (Comb2) # number of entries 7 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -180,11 +180,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -a0677317 14745600 0.000000e+00 4.292375e+04 3.070461e+03 2.704196e+06 1.166682e+11 63 -d48836c9 8294400 0.000000e+00 1.889580e+04 1.888554e+03 1.473872e+06 2.812818e+10 78 -69439308 774400 0.000000e+00 1.554382e+03 1.710542e+02 5.595775e+04 8.803307e+07 36 -5c532f3b 25600 0.000000e+00 7.450665e+01 4.146608e+00 4.097866e+03 3.062640e+05 55 -901ddf46 3686400 0.000000e+00 6.920145e+03 2.464678e+02 2.214446e+05 1.534373e+09 32 -3fc0fdcd 409600 0.000000e+00 1.036386e+03 1.971561e+02 6.011040e+04 6.455208e+07 58 3442589a 6400 0.000000e+00 1.971042e+01 5.545559e+00 3.744980e+02 7.965824e+03 19 +3fc0fdcd 409600 0.000000e+00 1.036386e+03 1.971561e+02 6.011040e+04 6.455208e+07 58 +901ddf46 3686400 0.000000e+00 6.718118e+03 9.120895e+02 1.907945e+06 1.305406e+10 284 +5c532f3b 25600 0.000000e+00 7.450665e+01 4.146608e+00 4.097866e+03 3.062640e+05 55 +69439308 774400 0.000000e+00 1.554382e+03 1.710542e+02 5.595775e+04 8.803307e+07 36 +d48836c9 8294400 0.000000e+00 1.889580e+04 1.888554e+03 1.473872e+06 2.812818e+10 78 +a0677317 14745600 0.000000e+00 4.292375e+04 3.070461e+03 2.704196e+06 1.166682e+11 63 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/ssyrk.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/ssyrk.sirocco index 1eaf9507c1c58c38871e877df6499a3450537dfb..44cd7d56b5f03a2b3e8672e21d7eec666753d042 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/ssyrk.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/ssyrk.sirocco @@ -7,7 +7,7 @@ # number of combinations 5 #################### -# COMB_1 +# COMB_0 # number of types devices 1 #################### @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cuda0_impl0 (Comb1) +# Model for cuda0_impl0 (Comb0) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,14 +36,14 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 3.885487e+03 3.138453e+02 1.266669e+06 4.953736e+09 326 -50116ceb 16588800 0.000000e+00 1.989457e+03 1.846987e+02 4.018704e+05 8.063950e+08 202 -a1187f60 7372800 0.000000e+00 6.844636e+02 6.633616e+01 2.457224e+05 1.697678e+08 359 +06ab5620 51200 0.000000e+00 6.011745e+01 1.013925e+01 1.322584e+03 8.177208e+04 22 34b59a6f 1548800 0.000000e+00 1.274568e+02 1.956594e+01 3.097201e+04 4.040621e+06 243 -6ab5620 51200 0.000000e+00 6.011745e+01 1.013925e+01 1.322584e+03 8.177208e+04 22 +a1187f60 7372800 0.000000e+00 6.861557e+02 5.462498e+01 5.338291e+05 3.686114e+08 778 +50116ceb 16588800 0.000000e+00 1.989457e+03 1.846987e+02 4.018704e+05 8.063950e+08 202 +ae9b1fd5 29491200 0.000000e+00 3.885487e+03 3.138453e+02 1.266669e+06 4.953736e+09 326 #################### -# COMB_0 +# COMB_2 # number of types devices 1 #################### @@ -62,7 +62,7 @@ a1187f60 7372800 0.000000e+00 6.844636e+02 6.633616e+01 2.45722 # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb0) +# Model for cpu0_impl0 (Comb2) # number of entries 7 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -72,16 +72,16 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 1.358370e+05 2.496094e+04 3.762684e+08 5.283701e+13 2770 -50116ceb 16588800 0.000000e+00 5.960685e+04 1.152639e+04 8.929106e+07 5.521380e+12 1498 -a1187f60 7372800 0.000000e+00 1.879259e+04 3.534541e+03 7.201320e+07 1.401187e+12 3832 -34b59a6f 1548800 0.000000e+00 2.022561e+03 3.698244e+02 8.274299e+06 1.729480e+10 4091 -6ab5620 51200 0.000000e+00 2.572545e+01 4.086938e+00 1.642570e+05 4.332233e+06 6385 -e78115ba 57600 0.000000e+00 4.722764e+01 8.558222e+00 1.558512e+03 7.602186e+04 33 72838077 819200 0.000000e+00 1.400813e+03 3.187817e+02 8.264794e+04 1.217700e+08 59 +e78115ba 57600 0.000000e+00 4.722764e+01 8.558222e+00 1.558512e+03 7.602186e+04 33 +06ab5620 51200 0.000000e+00 2.572545e+01 4.086938e+00 1.642570e+05 4.332233e+06 6385 +34b59a6f 1548800 0.000000e+00 2.022561e+03 3.698244e+02 8.274299e+06 1.729480e+10 4091 +a1187f60 7372800 0.000000e+00 1.843186e+04 3.887728e+03 1.452246e+08 2.795846e+12 7879 +50116ceb 16588800 0.000000e+00 5.960685e+04 1.152639e+04 8.929106e+07 5.521380e+12 1498 +ae9b1fd5 29491200 0.000000e+00 1.358370e+05 2.496094e+04 3.762684e+08 5.283701e+13 2770 #################### -# COMB_2 +# COMB_1 # number of types devices 1 #################### @@ -100,7 +100,7 @@ e78115ba 57600 0.000000e+00 4.722764e+01 8.558222e+00 1.55851 # number of implementations 1 ##### -# Model for cuda1_impl0 (Comb2) +# Model for cuda1_impl0 (Comb1) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -110,11 +110,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 3.928459e+03 3.071670e+02 1.312105e+06 5.186065e+09 334 -50116ceb 16588800 0.000000e+00 1.978866e+03 1.664848e+02 5.065897e+05 1.009569e+09 256 -a1187f60 7372800 0.000000e+00 6.835967e+02 5.317005e+01 1.893563e+05 1.302264e+08 277 +06ab5620 51200 0.000000e+00 6.087710e+01 1.044897e+01 6.087710e+02 3.815202e+04 10 34b59a6f 1548800 0.000000e+00 1.295589e+02 1.835748e+01 3.588782e+04 4.742936e+06 277 -6ab5620 51200 0.000000e+00 6.087710e+01 1.044897e+01 6.087710e+02 3.815202e+04 10 +a1187f60 7372800 0.000000e+00 6.879795e+02 4.496460e+01 4.189795e+05 2.894806e+08 609 +50116ceb 16588800 0.000000e+00 1.978866e+03 1.664848e+02 5.065897e+05 1.009569e+09 256 +ae9b1fd5 29491200 0.000000e+00 3.928459e+03 3.071670e+02 1.312105e+06 5.186065e+09 334 #################### # COMB_3 @@ -146,11 +146,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 3.795352e+03 3.099710e+02 1.335964e+06 5.104273e+09 352 -50116ceb 16588800 0.000000e+00 1.949808e+03 1.585039e+02 4.952512e+05 9.720259e+08 254 -a1187f60 7372800 0.000000e+00 6.738891e+02 5.051417e+01 2.459695e+05 1.666876e+08 365 +06ab5620 51200 0.000000e+00 5.816600e+01 9.053481e+00 1.105154e+03 6.583973e+04 19 34b59a6f 1548800 0.000000e+00 1.249379e+02 2.159779e+01 3.198410e+04 4.115441e+06 256 -6ab5620 51200 0.000000e+00 5.816600e+01 9.053481e+00 1.105154e+03 6.583973e+04 19 +a1187f60 7372800 0.000000e+00 6.738891e+02 5.051417e+01 2.459695e+05 1.666876e+08 365 +50116ceb 16588800 0.000000e+00 1.949808e+03 1.585039e+02 4.952512e+05 9.720259e+08 254 +ae9b1fd5 29491200 0.000000e+00 3.795352e+03 3.099710e+02 1.335964e+06 5.104273e+09 352 #################### # COMB_4 @@ -182,9 +182,9 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 3.938373e+03 2.754392e+02 1.606856e+06 6.359353e+09 408 -50116ceb 16588800 0.000000e+00 1.981274e+03 1.832435e+02 4.913559e+05 9.818378e+08 248 -a1187f60 7372800 0.000000e+00 6.796234e+02 6.194853e+01 2.113629e+05 1.448407e+08 311 +06ab5620 51200 0.000000e+00 6.400964e+01 1.458943e+01 8.961350e+02 6.034120e+04 14 34b59a6f 1548800 0.000000e+00 1.301442e+02 2.060864e+01 2.915229e+04 3.889137e+06 224 -6ab5620 51200 0.000000e+00 6.400964e+01 1.458943e+01 8.961350e+02 6.034120e+04 14 +a1187f60 7372800 0.000000e+00 6.796234e+02 6.194853e+01 2.113629e+05 1.448407e+08 311 +50116ceb 16588800 0.000000e+00 1.981274e+03 1.832435e+02 4.913559e+05 9.818378e+08 248 +ae9b1fd5 29491200 0.000000e+00 3.938373e+03 2.754392e+02 1.606856e+06 6.359353e+09 408 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/strmm.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/strmm.sirocco new file mode 100644 index 0000000000000000000000000000000000000000..cd63e6f59a5706abb01b8643137272de9e414c37 --- /dev/null +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/strmm.sirocco @@ -0,0 +1,104 @@ +################## +# Performance Model Version +45 + +#################### +# COMBs +# number of combinations +3 +#################### +# COMB_2 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +0 +#################### +# DEV_0 +# device id +0 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cpu0_impl0 (Comb2) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +a1187f60 7372800 0.000000e+00 1.759037e+04 3.255439e+03 4.012362e+07 7.299630e+11 2281 + +#################### +# COMB_0 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +1 +#################### +# DEV_0 +# device id +0 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cuda0_impl0 (Comb0) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +a1187f60 7372800 0.000000e+00 1.233237e+03 5.277861e+01 2.318485e+05 2.864478e+08 188 + +#################### +# COMB_1 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +1 +#################### +# DEV_0 +# device id +1 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cuda1_impl0 (Comb1) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +a1187f60 7372800 0.000000e+00 1.210116e+03 5.077157e+01 2.964784e+05 3.594048e+08 245 + diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/strsm.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/strsm.sirocco index 1a5783eb39d324db8c1dd064b4394eeaf71a51dd..36a65251bb39d6db379a3fc7a88b7dfc52325964 100644 --- a/simucore/perfmodels/.starpu/sampling/codelets/45/strsm.sirocco +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/strsm.sirocco @@ -7,7 +7,7 @@ # number of combinations 5 #################### -# COMB_0 +# COMB_2 # number of types devices 1 #################### @@ -26,7 +26,7 @@ # number of implementations 1 ##### -# Model for cpu0_impl0 (Comb0) +# Model for cpu0_impl0 (Comb2) # number of entries 7 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -36,13 +36,13 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 1.262354e+05 2.290170e+04 3.312416e+08 4.319067e+13 2624 -50116ceb 16588800 0.000000e+00 5.337615e+04 8.326371e+03 7.563401e+07 4.135291e+12 1417 -a1187f60 7372800 0.000000e+00 1.770446e+04 3.146495e+03 6.559501e+07 1.198005e+12 3705 -34b59a6f 1548800 0.000000e+00 1.942615e+03 3.132030e+02 9.050645e+06 1.803895e+10 4659 -6ab5620 51200 0.000000e+00 3.124193e+01 7.098676e+00 1.885451e+05 6.194624e+06 6035 -22f3afbe 460800 0.000000e+00 1.461281e+02 1.995835e+01 5.114484e+03 7.613117e+05 35 72838077 819200 0.000000e+00 2.006172e+03 3.121162e+02 1.263888e+05 2.596950e+08 63 +22f3afbe 460800 0.000000e+00 1.461281e+02 1.995835e+01 5.114484e+03 7.613117e+05 35 +06ab5620 51200 0.000000e+00 3.124193e+01 7.098676e+00 1.885451e+05 6.194624e+06 6035 +34b59a6f 1548800 0.000000e+00 1.942615e+03 3.132030e+02 9.050645e+06 1.803895e+10 4659 +a1187f60 7372800 0.000000e+00 1.891806e+04 3.953160e+03 1.392937e+08 2.750232e+12 7363 +50116ceb 16588800 0.000000e+00 5.337615e+04 8.326371e+03 7.563401e+07 4.135291e+12 1417 +ae9b1fd5 29491200 0.000000e+00 1.262354e+05 2.290170e+04 3.312416e+08 4.319067e+13 2624 #################### # COMB_4 @@ -74,14 +74,14 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 1.305248e+04 1.334385e+03 5.207939e+06 6.868697e+10 399 -50116ceb 16588800 0.000000e+00 6.227569e+03 8.472811e+02 1.955457e+06 1.240316e+10 314 -a1187f60 7372800 0.000000e+00 2.363204e+03 3.410126e+02 1.072895e+06 2.588265e+09 454 +06ab5620 51200 0.000000e+00 1.189339e+02 2.173609e+01 2.497612e+03 3.069724e+05 21 34b59a6f 1548800 0.000000e+00 9.588087e+02 1.079624e+02 1.716268e+05 1.666436e+08 179 -6ab5620 51200 0.000000e+00 1.189339e+02 2.173609e+01 2.497612e+03 3.069724e+05 21 +a1187f60 7372800 0.000000e+00 2.363204e+03 3.410126e+02 1.072895e+06 2.588265e+09 454 +50116ceb 16588800 0.000000e+00 6.227569e+03 8.472811e+02 1.955457e+06 1.240316e+10 314 +ae9b1fd5 29491200 0.000000e+00 1.305248e+04 1.334385e+03 5.207939e+06 6.868697e+10 399 #################### -# COMB_2 +# COMB_1 # number of types devices 1 #################### @@ -100,7 +100,7 @@ a1187f60 7372800 0.000000e+00 2.363204e+03 3.410126e+02 1.07289 # number of implementations 1 ##### -# Model for cuda1_impl0 (Comb2) +# Model for cuda1_impl0 (Comb1) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -110,11 +110,11 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 1.285236e+04 1.196289e+03 4.961010e+06 6.431309e+10 386 -50116ceb 16588800 0.000000e+00 5.951177e+03 5.675201e+02 1.690134e+06 1.014976e+10 284 -a1187f60 7372800 0.000000e+00 2.336819e+03 2.871770e+02 9.136963e+05 2.167389e+09 391 +06ab5620 51200 0.000000e+00 6.696491e+01 1.309227e+01 2.209842e+03 1.536383e+05 33 34b59a6f 1548800 0.000000e+00 9.794784e+02 1.067792e+02 1.557371e+05 1.543540e+08 159 -6ab5620 51200 0.000000e+00 6.696491e+01 1.309227e+01 2.209842e+03 1.536383e+05 33 +a1187f60 7372800 0.000000e+00 1.946299e+03 2.988993e+02 3.044012e+06 6.064287e+09 1564 +50116ceb 16588800 0.000000e+00 5.951177e+03 5.675201e+02 1.690134e+06 1.014976e+10 284 +ae9b1fd5 29491200 0.000000e+00 1.285236e+04 1.196289e+03 4.961010e+06 6.431309e+10 386 #################### # COMB_3 @@ -146,14 +146,14 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 1.284788e+04 1.232936e+03 6.231224e+06 8.079530e+10 485 -50116ceb 16588800 0.000000e+00 5.870740e+03 6.354548e+02 1.943215e+06 1.154177e+10 331 -a1187f60 7372800 0.000000e+00 2.366672e+03 3.276943e+02 8.401685e+05 2.026524e+09 355 +06ab5620 51200 0.000000e+00 1.527756e+02 3.246283e+01 2.597185e+03 4.147017e+05 17 34b59a6f 1548800 0.000000e+00 9.462385e+02 1.249538e+02 1.750541e+05 1.685314e+08 185 -6ab5620 51200 0.000000e+00 1.527756e+02 3.246283e+01 2.597185e+03 4.147017e+05 17 +a1187f60 7372800 0.000000e+00 2.366672e+03 3.276943e+02 8.401685e+05 2.026524e+09 355 +50116ceb 16588800 0.000000e+00 5.870740e+03 6.354548e+02 1.943215e+06 1.154177e+10 331 +ae9b1fd5 29491200 0.000000e+00 1.284788e+04 1.232936e+03 6.231224e+06 8.079530e+10 485 #################### -# COMB_1 +# COMB_0 # number of types devices 1 #################### @@ -172,7 +172,7 @@ a1187f60 7372800 0.000000e+00 2.366672e+03 3.276943e+02 8.40168 # number of implementations 1 ##### -# Model for cuda0_impl0 (Comb1) +# Model for cuda0_impl0 (Comb0) # number of entries 5 # sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx @@ -182,9 +182,9 @@ nan nan nan # not multiple-regression-base 0 # hash size flops mean (us) dev (us) sum sum2 n -ae9b1fd5 29491200 0.000000e+00 1.272736e+04 1.188134e+03 5.409128e+06 6.944388e+10 425 -50116ceb 16588800 0.000000e+00 5.985624e+03 7.297118e+02 1.675975e+06 1.018085e+10 280 -a1187f60 7372800 0.000000e+00 2.310110e+03 2.968759e+02 1.016448e+06 2.386886e+09 440 +06ab5620 51200 0.000000e+00 1.183823e+02 2.064199e+01 1.775735e+03 2.166070e+05 15 34b59a6f 1548800 0.000000e+00 9.767572e+02 1.265878e+02 1.670255e+05 1.658835e+08 171 -6ab5620 51200 0.000000e+00 1.183823e+02 2.064199e+01 1.775735e+03 2.166070e+05 15 +a1187f60 7372800 0.000000e+00 1.974544e+03 2.968629e+02 3.293539e+06 6.650235e+09 1668 +50116ceb 16588800 0.000000e+00 5.985624e+03 7.297118e+02 1.675975e+06 1.018085e+10 280 +ae9b1fd5 29491200 0.000000e+00 1.272736e+04 1.188134e+03 5.409128e+06 6.944388e+10 425 diff --git a/simucore/perfmodels/.starpu/sampling/codelets/45/strtri.sirocco b/simucore/perfmodels/.starpu/sampling/codelets/45/strtri.sirocco new file mode 100644 index 0000000000000000000000000000000000000000..e7d9c50b49fd086e454feea6680b4f1da106ba0f --- /dev/null +++ b/simucore/perfmodels/.starpu/sampling/codelets/45/strtri.sirocco @@ -0,0 +1,40 @@ +################## +# Performance Model Version +45 + +#################### +# COMBs +# number of combinations +1 +#################### +# COMB_2 +# number of types devices +1 +#################### +# DEV_0 +# device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3, MPI_MS - 5) +0 +#################### +# DEV_0 +# device id +0 +#################### +# DEV_0 +# number of cores +1 +########## +# number of implementations +1 +##### +# Model for cpu0_impl0 (Comb2) +# number of entries +1 +# sumlnx sumlnx2 sumlny sumlnxlny alpha beta n minx maxx +0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 nan nan 0 0 0 +# a b c +nan nan nan +# not multiple-regression-base +0 +# hash size flops mean (us) dev (us) sum sum2 n +901ddf46 3686400 0.000000e+00 1.206892e+04 2.330425e+03 2.896541e+06 3.626154e+10 240 +