1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113
|
#!/bin/bash
# Bash script to submit many files to Cori/Edison/Queue
#BSUB -P CSC289
#BSUB -W 2:00
#BSUB -nnodes 45
#BSUB -alloc_flags gpumps
#BSUB -J superlu_gpu
EXIT_SUCCESS=0
EXIT_HOST=1
EXIT_PARAM=2
module load essl
module load cmake/3.11.3
module load cuda/10.1.168
CUR_DIR=`pwd`
FILE_DIR=$CUR_DIR/EXAMPLE
INPUT_DIR=$MEMBERWORK/csc289/matrix/HTS
# INPUT_DIR=$CUR_DIR/../EXAMPLE
FILE_NAME=pddrive
FILE=$FILE_DIR/$FILE_NAME
nprows=( 1 )
npcols=( 1 )
for ((i = 0; i < ${#npcols[@]}; i++)); do
NROW=${nprows[i]}
NCOL=${npcols[i]}
# NROW=36
CORE_VAL=`expr $NCOL \* $NROW`
RANK_PER_RS=1
GPU_PER_RANK=0
#PARTITION=debug
PARTITION=regular
LICENSE=SCRATCH
TIME=00:20:00
if [[ $NERSC_HOST == edison ]]
then
CONSTRAINT=0
fi
if [[ $NERSC_HOST == cori ]]
then
CONSTRAINT=haswell
fi
for GPU_PER_RANK in 1
do
for NTH in 7
do
RS_VAL=`expr $CORE_VAL / $RANK_PER_RS`
MOD_VAL=`expr $CORE_VAL % $RANK_PER_RS`
if [[ $MOD_VAL -ne 0 ]]
then
RS_VAL=`expr $RS_VAL + 1`
fi
OMP_NUM_THREADS=$NTH
TH_PER_RS=`expr $NTH \* $RANK_PER_RS`
GPU_PER_RS=`expr $RANK_PER_RS \* $GPU_PER_RANK`
# for MAT in copter2.mtx
# for MAT in rajat16.mtx
# for MAT in ExaSGD/118_1536/globalmat.datnh
# for MAT in copter2.mtx gas_sensor.mtx matrix-new_3.mtx xenon2.mtx shipsec1.mtx xenon1.mtx g7jac160.mtx g7jac140sc.mtx mark3jac100sc.mtx ct20stif.mtx vanbody.mtx ncvxbqp1.mtx dawson5.mtx 2D_54019_highK.mtx gridgena.mtx epb3.mtx torso2.mtx torsion1.mtx boyd1.bin hvdc2.mtx rajat16.mtx hcircuit.mtx
# for MAT in copter2.mtx gas_sensor.mtx matrix-new_3.mtx av41092.mtx xenon2.mtx c-71.mtx shipsec1.mtx xenon1.mtx g7jac160.mtx g7jac140sc.mtx mark3jac100sc.mtx ct20stif.mtx vanbody.mtx ncvxbqp1.mtx dawson5.mtx c-59.mtx 2D_54019_highK.mtx gridgena.mtx epb3.mtx torso2.mtx finan512.mtx twotone.mtx torsion1.mtx jan99jac120.mtx boyd1.mtx c-73b.mtx hvdc2.mtx rajat16.mtx hcircuit.mtx
# for MAT in matrix05_ntime=2/s1_mat_0_126936.bin
# for MAT in matrix05_ntime=2/s1_mat_0_126936.bin A30_P0/A30_015_0_25356.bin
# for MAT in matrix05_ntime=2/s1_mat_0_126936.bin
# for MAT in ../mathias/Li4244.bin ../mathias/DG_GrapheneDisorder_32768.bin
#for MAT in copter2.mtx epb3.mtx gridgena.mtx vanbody.mtx shipsec1.mtx dawson5.mtx gas_sensor.mtx rajat16.mtx
#for MAT in rajat16.mtx
#for MAT in epb3.mtx
#for MAT in ../globalmat118_1536.bin
#for MAT in ../mat_it_001_sml.mtx ../mat_it_001_med.mtx
#for MAT in ../mat_it_001_med.mtx
#for MAT in ../s1_mat_0_126936.bin
for MAT in copter2.mtx
#for MAT in ../full_1000.rua
# for MAT in A30_P0/A30_015_0_25356.bin
# for MAT in A64/A64_001_0_1204992.bin
#for MAT in ../big.rua
#for MAT in ../g4.rua
# for MAT in StocF-1465.bin
# for MAT in /mathias/DG_GrapheneDisorder_8192.bin /mathias/DNA_715_64cell.bin /mathias/LU_C_BN_C_4by2.bin /mathias/Li4244.bin
# for MAT in HTS/copter2.mtx
# for MAT in LU_C_BN_C_4by2.bin Li4244.bin atmosmodj.bin Ga19As19H42.bin Geo_1438.bin StocF-1465.bin
do
export OMP_NUM_THREADS=$OMP_NUM_THREADS
mkdir -p ${MAT}_summit
echo "jsrun -n $RS_VAL -a $RANK_PER_RS -c $TH_PER_RS -g $GPU_PER_RS -b packed:$NTH $FILE -c $NCOL -r $NROW $INPUT_DIR/$MAT | tee ./${MAT}_summit_new_LU/SLU.o_mpi_${NROW}x${NCOL}_OMP_${OMP_NUM_THREADS}_GPU_${GPU_PER_RANK}"
jsrun -n $RS_VAL -a $RANK_PER_RS -c $TH_PER_RS -g $GPU_PER_RS -b packed:$NTH '--smpiargs=-x PAMI_DISABLE_CUDA_HOOK=1 -disable_gpu_hooks' nvprof --profile-from-start off $FILE -c $NCOL -r $NROW $INPUT_DIR/$MAT | tee ./${MAT}_summit/SLU.o_mpi_${NROW}x${NCOL}_OMP_${OMP_NUM_THREADS}_GPU_${GPU_PER_RANK}
# jsrun -n $RS_VAL -a $RANK_PER_RS -c $TH_PER_RS -g $GPU_PER_RS -b packed:$NTH '--smpiargs=-x PAMI_DISABLE_CUDA_HOOK=1 -disable_gpu_hooks' nvprof --profile-from-start off --print-gpu-trace --print-api-trace --kernels "dlsum_fmod_inv_cuda" --metrics "achieved_occupancy" $FILE -c $NCOL -r $NROW $INPUT_DIR/$MAT | tee ./${MAT}_summit/SLU.o_mpi_${NROW}x${NCOL}_OMP_${OMP_NUM_THREADS}_GPU_${GPU_PER_RANK}
# jsrun -n $RS_VAL -a $RANK_PER_RS -c $TH_PER_RS -g $GPU_PER_RS -b packed:$NTH '--smpiargs=-x PAMI_DISABLE_CUDA_HOOK=1 -disable_gpu_hooks' nvprof $FILE -c $NCOL -r $NROW $INPUT_DIR/$MAT | tee ./${MAT}_summit/SLU.o_mpi_${NROW}x${NCOL}_OMP_${OMP_NUM_THREADS}_GPU_${GPU_PER_RANK}
done
#one
done
done
done
exit $EXIT_SUCCESS
|