StarPU MORSE Nightly Tests

Table of Contents

Note

  • dgemm, sgemm and spotrf miriel benchmarks are run on OPA miriel nodes.
  • other spotrf benchmarks are run on sirocco nodes from 2017, december 15th (sequential benchmarks used to be run on mirage nodes).

Tests for master #HEAD with profile spotrf_miriel_mpi

Execution on queue routage with 4 node(s) (-C "miriel&omnipath") of the following script:

printenv | grep SLURM 
printenv | grep STARPU_ 
module list 
ulimit -c 0 
export STARPU_WORKER_STATS=1 
export STARPU_CALIBRATE=2 
export STARPU_LIMIT_CPU_MEM=120000 
export STARPU_LIMIT_MIN_SUBMITTED_TASKS=15000 
export STARPU_LIMIT_MAX_SUBMITTED_TASKS=16000 
export p=$(perl -e 'print int(sqrt('$SLURM_NNODES')), "n"') 
procs=$(echo $SLURM_JOB_CPUS_PER_NODE | awk -F'(' '{print $1}') 
export nbthreads=$(($procs - 2)) 
export niter=${NITER:-10} 
mpiexec -np $SLURM_NNODES --map-by node hostname 
mpiexec --mca btl self,openib -np $SLURM_NNODES --map-by node -- $PWD/testing/chameleon_stesting --op \ 
   potrf --check --niter $niter --ib 64 --nb 320 --n 64000:64000:64000 --P $p --threads $nbthreads 
date slurm morse_scm starpu_scm output Id Function threads gpus P Q mtxfmt nb uplo n lda seedA time gflops RETURN deviation
2024-04-25 slurm 0ac42d5 100cb86 output - - - - - - - - - - - - - - - -
2024-04-24 slurm 0ac42d5 100cb86 output - - - - - - - - - - - - - - - -
2024-04-23 slurm 0ac42d5 ae032cb output - - - - - - - - - - - - - - - -
2024-04-22 slurm 0ac42d5 af2178d output 3 spotrf 22 0 2 2 0 320 121 64000 64000 1957747793 1.036040e+02 8.434363e+02 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 2 spotrf 22 0 2 2 0 320 121 64000 64000 1714636915 7.896435e+01 1.106618e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 1 spotrf 22 0 2 2 0 320 121 64000 64000 1681692777 9.413865e+01 9.282412e+02 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 0 spotrf 22 0 2 2 0 320 121 64000 64000 846930886 8.815299e+01 9.912696e+02 SUCCESS 0
2024-04-21 slurm 0ac42d5 af2178d output - - - - - - - - - - - - - - - -
2024-04-20 slurm 0ac42d5 af2178d output 4 spotrf 22 0 2 2 0 320 121 64000 64000 424238335 9.865747e+01 8.857249e+02 SUCCESS 0
2024-04-20 slurm 0ac42d5 af2178d output 3 spotrf 22 0 2 2 0 320 121 64000 64000 1957747793 8.184568e+01 1.067660e+03 SUCCESS 0
2024-04-20 slurm 0ac42d5 af2178d output 2 spotrf 22 0 2 2 0 320 121 64000 64000 1714636915 1.082630e+02 8.071399e+02 SUCCESS 0
2024-04-20 slurm 0ac42d5 af2178d output 1 spotrf 22 0 2 2 0 320 121 64000 64000 1681692777 9.946011e+01 8.785772e+02 SUCCESS 0
2024-04-20 slurm 0ac42d5 af2178d output 0 spotrf 22 0 2 2 0 320 121 64000 64000 846930886 1.025910e+02 8.517647e+02 SUCCESS 0
2024-04-19 slurm 0ac42d5 5858752 output 3 spotrf 22 0 2 2 0 320 121 64000 64000 1957747793 8.779900e+01 9.952662e+02 SUCCESS 0
2024-04-19 slurm 0ac42d5 5858752 output 2 spotrf 22 0 2 2 0 320 121 64000 64000 1714636915 8.027879e+01 1.088499e+03 SUCCESS 0
2024-04-19 slurm 0ac42d5 5858752 output 1 spotrf 22 0 2 2 0 320 121 64000 64000 1681692777 7.716397e+01 1.132438e+03 SUCCESS 0
2024-04-19 slurm 0ac42d5 5858752 output 0 spotrf 22 0 2 2 0 320 121 64000 64000 846930886 8.606599e+01 1.015307e+03 SUCCESS 0
2024-04-18 slurm 0ac42d5 6813015 output - - - - - - - - - - - - - - - -
2024-04-12 slurm 8d2fdd4 dcfc8b4 output - - - - - - - - - - - - - - - -
2024-04-11 slurm 8d2fdd4 33fa9a7 output - - - - - - - - - - - - - - - -

See also the test archives: 2019 2020 2021 2022 2023 2024 .

Tests for master #HEAD with profile spotrf_miriel_mpi_nmad

Execution on queue routage with 4 node(s) (-C "miriel&omnipath") of the following script:

printenv | grep SLURM 
printenv | grep STARPU_ 
module list 
ulimit -c 0 
export STARPU_WORKER_STATS=1 
export STARPU_CALIBRATE=2 
export STARPU_LIMIT_CPU_MEM=120000 
export STARPU_LIMIT_MIN_SUBMITTED_TASKS=15000 
export STARPU_LIMIT_MAX_SUBMITTED_TASKS=16000 
export p=$(perl -e 'print int(sqrt('$SLURM_NNODES')), "n"') 
procs=$(echo $SLURM_JOB_CPUS_PER_NODE | awk -F'(' '{print $1}') 
export nbthreads=$(($procs - 2)) 
type mpiexec 
export niter=${NITER:-10} 
mpiexec -DPATH=${PATH} -DLD_LIBRARY_PATH=${LD_LIBRARY_PATH} -np $SLURM_NNODES hostname 
mpiexec -DPATH=${PATH} -DLD_LIBRARY_PATH=${LD_LIBRARY_PATH} -np $SLURM_NNODES $PWD/testing/chameleon_stesting \ 
   --op potrf --check --niter $niter --ib 64 --nb 320 --n 64000:64000:64000 --P $p --threads $nbthreads 
date slurm morse_scm starpu_scm output Id Function threads gpus P Q mtxfmt nb uplo n lda seedA time gflops RETURN deviation
2024-04-25 slurm 0ac42d5 100cb86 output - - - - - - - - - - - - - - - -
2024-04-24 slurm 0ac42d5 100cb86 output - - - - - - - - - - - - - - - -
2024-04-23 slurm 0ac42d5 ae032cb output - - - - - - - - - - - - - - - -
2024-04-22 slurm 0ac42d5 af2178d output 9 spotrf 22 0 2 2 0 320 121 64000 64000 465755779 1.672272e+01 5.225429e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 8 spotrf 22 0 2 2 0 320 121 64000 64000 629974776 1.676001e+01 5.213801e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 7 spotrf 22 0 2 2 0 320 121 64000 64000 1440703202 1.677838e+01 5.208094e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 6 spotrf 22 0 2 2 0 320 121 64000 64000 463880093 1.681005e+01 5.198281e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 5 spotrf 22 0 2 2 0 320 121 64000 64000 697820186 1.676407e+01 5.212539e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 4 spotrf 22 0 2 2 0 320 121 64000 64000 59509029 1.681523e+01 5.196679e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 3 spotrf 22 0 2 2 0 320 121 64000 64000 1020068712 1.676489e+01 5.212286e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 2 spotrf 22 0 2 2 0 320 121 64000 64000 1071003750 1.677363e+01 5.209570e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 1 spotrf 22 0 2 2 0 320 121 64000 64000 1108039844 1.682640e+01 5.193231e+03 SUCCESS 0
2024-04-22 slurm 0ac42d5 af2178d output 0 spotrf 22 0 2 2 0 320 121 64000 64000 1281361282 1.708423e+01 5.114857e+03 SUCCESS 0
2024-04-21 slurm 0ac42d5 af2178d output - - - - - - - - - - - - - - - -
2024-04-20 slurm 0ac42d5 af2178d output 9 spotrf 22 0 2 2 0 320 121 64000 64000 890133021 1.668205e+01 5.238168e+03 SUCCESS 0
2024-04-20 slurm 0ac42d5 af2178d output 8 spotrf 22 0 2 2 0 320 121 64000 64000 1588110562 1.669560e+01 5.233917e+03 SUCCESS 0
2024-04-20 slurm 0ac42d5 af2178d output 7 spotrf 22 0 2 2 0 320 121 64000 64000 1128502328 1.669529e+01 5.234013e+03 SUCCESS 0
2024-04-20 slurm 0ac42d5 af2178d output 6 spotrf 22 0 2 2 0 320 121 64000 64000 2123414232 1.671866e+01 5.226696e+03 SUCCESS 0
2024-04-20 slurm 0ac42d5 af2178d output 5 spotrf 22 0 2 2 0 320 121 64000 64000 838576346 1.667026e+01 5.241872e+03 SUCCESS 0
2024-04-20 slurm 0ac42d5 af2178d output 4 spotrf 22 0 2 2 0 320 121 64000 64000 1087043833 1.670585e+01 5.230704e+03 SUCCESS 0

See also the test archives: 2020 2021 2022 2023 2024 .

Tests for master #HEAD with profile spotrf_miriel_mpi_mad_mpi

Execution on queue routage with 4 node(s) (-C "miriel&omnipath") of the following script:

printenv | grep SLURM 
printenv | grep STARPU_ 
module list 
ulimit -c 0 
export STARPU_WORKER_STATS=1 
export STARPU_CALIBRATE=2 
export STARPU_LIMIT_CPU_MEM=120000 
export STARPU_LIMIT_MIN_SUBMITTED_TASKS=15000 
export STARPU_LIMIT_MAX_SUBMITTED_TASKS=16000 
export p=$(perl -e 'print int(sqrt('$SLURM_NNODES')), "n"') 
procs=$(echo $SLURM_JOB_CPUS_PER_NODE | awk -F'(' '{print $1}') 
export nbthreads=$(($procs - 2)) 
type mpiexec 
export niter=${NITER:-10} 
mpiexec -DPATH=${PATH} -DLD_LIBRARY_PATH=${LD_LIBRARY_PATH} -np $SLURM_NNODES hostname 
mpiexec -DPATH=${PATH} -DLD_LIBRARY_PATH=${LD_LIBRARY_PATH} -np $SLURM_NNODES $PWD/testing/chameleon_stesting \ 
   --op potrf --check --niter $niter --ib 64 --nb 320 --n 64000:64000:64000 --P $p --threads $nbthreads 
date slurm morse_scm starpu_scm output Id Function threads gpus P Q mtxfmt nb uplo n lda seedA time gflops RETURN deviation
2024-04-25 slurm 0ac42d5 100cb86 output - - - - - - - - - - - - - - - -
2024-04-24 slurm 0ac42d5 100cb86 output - - - - - - - - - - - - - - - -
2024-04-23 slurm 0ac42d5 ae032cb output - - - - - - - - - - - - - - - -
2024-04-22 slurm 0ac42d5 af2178d output - - - - - - - - - - - - - - - -
2024-04-21 slurm 0ac42d5 af2178d output - - - - - - - - - - - - - - - -
2024-04-20 slurm 0ac42d5 af2178d output - - - - - - - - - - - - - - - -
2024-04-19 slurm 0ac42d5 5858752 output - - - - - - - - - - - - - - - -
2024-04-18 slurm 0ac42d5 6813015 output - - - - - - - - - - - - - - - -
2024-04-12 slurm 8d2fdd4 dcfc8b4 output - - - - - - - - - - - - - - - -
2024-04-11 slurm 8d2fdd4 015e91b output - - - - - - - - - - - - - - - -
2024-04-10 slurm 8d2fdd4 b2b4f41 output - - - - - - - - - - - - - - - -
2024-04-09 slurm 8d2fdd4 f19f2ac output - - - - - - - - - - - - - - - -
2024-04-08 slurm 8d2fdd4 f19f2ac output - - - - - - - - - - - - - - - -
2024-04-07 slurm 8d2fdd4 f19f2ac output - - - - - - - - - - - - - - - -
2024-04-06 slurm 8d2fdd4 f19f2ac output - - - - - - - - - - - - - - - -
2024-04-05 slurm 8d2fdd4 3bc35bd output - - - - - - - - - - - - - - - -
2024-04-04 slurm 8d2fdd4 91d6206 output - - - - - - - - - - - - - - - -
2024-04-03 slurm af8d785 bb2b4a1 output - - - - - - - - - - - - - - - -
2024-04-02 slurm 30c5300 382f97a output - - - - - - - - - - - - - - - -
2024-04-01 slurm 30c5300 882aba6 output - - - - - - - - - - - - - - - -

See also the test archives: 2020 2021 2022 2023 2024 .

Tests for master #HEAD with profile spotrf_miriel_mpi_sim

Execution on queue routage with 1 node(s) (-C "miriel&omnipath") of the following script:

printenv | grep SLURM 
printenv | grep STARPU_ 
module list 
ulimit -c 0 
export STARPU_HOME="$prefix_dir/../morse/master/simucore/perfmodels/" 
export STARPU_HOSTNAME=sirocco 
export STARPU_CALIBRATE=0 
export STARPU_WORKER_STATS=1 
export STARPU_LIMIT_CPU_MEM=120000 
export STARPU_LIMIT_MIN_SUBMITTED_TASKS=15000 
export STARPU_LIMIT_MAX_SUBMITTED_TASKS=16000 
export STARPU_MALLOC_SIMULATION_FOLD=2 
SLURM_JOB_NUM_NODES=4 
SLURM_CPUS_ON_NODE=24 
p=$(perl -e 'print int(sqrt('$SLURM_JOB_NUM_NODES')), "n"') 
procs=$(echo $SLURM_JOB_CPUS_PER_NODE | awk -F'(' '{print $1}') 
export nbthreads=$(($procs - 1)) 
time $rootdir/starpu.inst/bin/starpu_smpirun -np $SLURM_JOB_NUM_NODES -platform $HOME/.starpu/cluster.xml \ 
   -hostfile $HOME/.starpu/hostfile --cfg=surf/precision:0.000010 --cfg=maxmin/precision:0.000010 --cfg=network/model:IB \ 
   --cfg=smpi/test:0.000010 ./testing/chameleon_stesting --op potrf --ib 64 --nb 320 --n 64000:64000:64000 \ 
   --P $p 
date slurm morse_scm starpu_scm output Id Function threads gpus P Q mtxfmt nb uplo n lda seedA time gflops RETURN deviation
2024-04-25 slurm 0ac42d5 100cb86 output - - - - - - - - - - - - - - - -
2024-04-24 slurm 0ac42d5 100cb86 output - - - - - - - - - - - - - - - -
2024-04-23 slurm 0ac42d5 ae032cb output - - - - - - - - - - - - - - - -
2024-04-22 slurm 0ac42d5 af2178d output - - - - - - - - - - - - - - - -
2024-04-21 slurm 0ac42d5 af2178d output - - - - - - - - - - - - - - - -
2024-04-20 slurm 0ac42d5 af2178d output - - - - - - - - - - - - - - - -
2024-04-19 slurm 0ac42d5 5858752 output - - - - - - - - - - - - - - - -
2024-04-18 slurm 0ac42d5 6813015 output - - - - - - - - - - - - - - - -
2024-04-12 slurm 8d2fdd4 dcfc8b4 output - - - - - - - - - - - - - - - -
2024-04-11 slurm 8d2fdd4 6b8ab2b output - - - - - - - - - - - - - - - -
2024-04-10 slurm 8d2fdd4 b2b4f41 output - - - - - - - - - - - - - - - -
2024-04-09 slurm 8d2fdd4 f19f2ac output - - - - - - - - - - - - - - - -
2024-04-08 slurm 8d2fdd4 f19f2ac output - - - - - - - - - - - - - - - -
2024-04-07 slurm 8d2fdd4 f19f2ac output - - - - - - - - - - - - - - - -
2024-04-06 slurm 8d2fdd4 f19f2ac output - - - - - - - - - - - - - - - -
2024-04-05 slurm 8d2fdd4 3bc35bd output - - - - - - - - - - - - - - - -
2024-04-04 slurm 8d2fdd4 91d6206 output - - - - - - - - - - - - - - - -
2024-04-03 slurm af8d785 bb2b4a1 output - - - - - - - - - - - - - - - -
2024-04-02 slurm 30c5300 382f97a output - - - - - - - - - - - - - - - -
2024-04-01 slurm 30c5300 882aba6 output - - - - - - - - - - - - - - - -

See also the test archives: 2020 2021 2022 2023 2024 .

Plot

results_master_HEAD_spotrf_miriel_mpi.png

Raw data

Author: gitlab-starpu

Created: 2024-04-25 Thu 21:27

Validate