guilhem_lavaux/JaxPM
0
0
Fork 1
mirror of https://github.com/DifferentiableUniverseInitiative/JaxPM.git synced 2025-04-08 04:40:53 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

merge with JZ

Browse source
This commit is contained in:
Wassim KABALAN 2024-10-18 14:59:40 -04:00
commit ab86699c88
7 changed files with 226 additions and 131 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

153
benchmarks/bench_pm.py
View file

@ -26,14 +26,17 @@ from jax.sharding import PartitionSpec as P
from jaxpm.kernels import interpolate_power_spectrum from jaxpm.kernels import interpolate_power_spectrum
from jaxpm.painting import cic_paint_dx from jaxpm.painting import cic_paint_dx
from jaxpm.pm import linear_field, lpt, make_ode_fn from jaxpm.pm import linear_field, lpt, make_ode_fn
from jax import make_jaxpr
def run_simulation(mesh_shape, def run_simulation(mesh_shape,
box_size, box_size,
halo_size, halo_size,
solver_choice, solver_choice,
iterations, iterations,
pdims=None): hlo_print,
trace,
pdims=None,
output_path="."):
@jax.jit @jax.jit
def simulate(omega_c, sigma8): def simulate(omega_c, sigma8):
@ -60,7 +63,6 @@ def run_simulation(mesh_shape,
solver = Tsit5() solver = Tsit5()
elif solver_choice == "lpt": elif solver_choice == "lpt":
lpt_field = cic_paint_dx(dx, halo_size=halo_size) lpt_field = cic_paint_dx(dx, halo_size=halo_size)
print(f"TYPE of lpt_field: {type(lpt_field)}")
return lpt_field, {"num_steps": 0} return lpt_field, {"num_steps": 0}
else: else:
raise ValueError( raise ValueError(
@ -92,6 +94,7 @@ def run_simulation(mesh_shape,
def run(): def run():
# Warm start # Warm start
<<<<<<< HEAD
chrono_fun = Timer() chrono_fun = Timer()
RangePush("warmup") RangePush("warmup")
final_field, stats = chrono_fun.chrono_jit(simulate, final_field, stats = chrono_fun.chrono_jit(simulate,
@ -108,6 +111,39 @@ def run_simulation(mesh_shape,
ndarray_arg=0) ndarray_arg=0)
RangePop() RangePop()
return final_field, stats, chrono_fun return final_field, stats, chrono_fun
=======
if hlo_print:
jaxpr = make_jaxpr(simulate)(0.32, 0.8)
lowered = jax.jit(simulate).lower(0.32, 0.8)
lower_as_text = lowered.as_text()
compiled = lowered.compile()
compiled_again = jax.jit(simulate).lower(0.32, 0.8).compile()
return jaxpr , compiled , compiled_again
elif trace:
jit_output = f"{output_path}/jit_trace"
first_run_output = f"{output_path}/first_run_trace"
second_run_output = f"{output_path}/second_run_trace"
with jax.profiler.trace(jit_output , create_perfetto_trace=True):
final_field, stats = simulate(0.32, 0.8)
final_field.block_until_ready()
with jax.profiler.trace(first_run_output , create_perfetto_trace=True):
final_field, stats = simulate(0.32, 0.8)
final_field.block_until_ready()
with jax.profiler.trace(second_run_output , create_perfetto_trace=True):
final_field, stats = simulate(0.32, 0.8)
final_field.block_until_ready()
else:
chrono_fun = Timer()
RangePush("warmup")
final_field, stats = chrono_fun.chrono_jit(simulate, 0.32, 0.8 , ndarray_arg = 0)
RangePop()
sync_global_devices("warmup")
for i in range(iterations):
RangePush(f"sim iter {i}")
final_field, stats = chrono_fun.chrono_fun(simulate, 0.32, 0.8 , ndarray_arg = 0)
RangePop()
return final_field, stats, chrono_fun
>>>>>>> glab/ASKabalan/jaxdecomp_proto
if jax.device_count() > 1: if jax.device_count() > 1:
devices = mesh_utils.create_device_mesh(pdims) devices = mesh_utils.create_device_mesh(pdims)
@ -151,7 +187,7 @@ if __name__ == "__main__":
'--halo_size', '--halo_size',
type=int, type=int,
help='Halo size', help='Halo size',
required=True) default=None)
parser.add_argument('-s', parser.add_argument('-s',
'--solver', '--solver',
type=str, type=str,
@ -161,12 +197,7 @@ if __name__ == "__main__":
"LeapfrogMidpoint", "leapfrogmidpoint", "lfm", "LeapfrogMidpoint", "leapfrogmidpoint", "lfm",
"lpt" "lpt"
], ],
required=True) default="lpt")
parser.add_argument('-i',
'--iterations',
type=int,
help='Number of iterations',
default=10)
parser.add_argument('-o', parser.add_argument('-o',
'--output_path', '--output_path',
type=str, type=str,
@ -181,15 +212,39 @@ if __name__ == "__main__":
type=int, type=int,
help='Number of nodes', help='Number of nodes',
default=1) default=1)
<<<<<<< HEAD
=======
parser.add_argument('-i',
'--iterations',
type=int,
help='Number of iterations',
default=10)
group = parser.add_mutually_exclusive_group()
group.add_argument('-hlo',
'--hlo_print',
action='store_true',
help='Print hlo generated by XLA')
group.add_argument('-t',
'--trace',
action='store_true',
help='Profile using tensorboard')
>>>>>>> glab/ASKabalan/jaxdecomp_proto
args = parser.parse_args() args = parser.parse_args()
mesh_size = args.mesh_size mesh_size = args.mesh_size
box_size = [args.box_size] * 3 box_size = [args.box_size] * 3
halo_size = args.halo_size halo_size = args.mesh_size // 8 if args.halo_size is None else args.halo_size
solver_choice = args.solver solver_choice = args.solver
iterations = args.iterations iterations = args.iterations
output_path = args.output_path output_path = args.output_path
os.makedirs(output_path, exist_ok=True) os.makedirs(output_path, exist_ok=True)
<<<<<<< HEAD
=======
hlo_print = args.hlo_print
trace = args.trace
nb_gpus = jax.device_count()
>>>>>>> glab/ASKabalan/jaxdecomp_proto
print(f"solver choice: {solver_choice}") print(f"solver choice: {solver_choice}")
match solver_choice: match solver_choice:
@ -213,9 +268,11 @@ if __name__ == "__main__":
if args.pdims: if args.pdims:
pdims = tuple(map(int, args.pdims.split("x"))) pdims = tuple(map(int, args.pdims.split("x")))
else: else:
pdims = (1, 1) pdims = (1, jax.device_count())
pdm_str = f"{pdims[0]}x{pdims[1]}"
mesh_shape = [mesh_size] * 3 mesh_shape = [mesh_size] * 3
<<<<<<< HEAD
final_field, stats, chrono_fun = run_simulation(mesh_shape, box_size, final_field, stats, chrono_fun = run_simulation(mesh_shape, box_size,
halo_size, solver_choice, halo_size, solver_choice,
@ -226,6 +283,50 @@ if __name__ == "__main__":
) )
metadata = { metadata = {
=======
if trace:
trace_folder = f"{output_path}/profiling/jaxpm/{nb_gpus}/{mesh_shape[0]}_{int(box_size[0])}/{pdm_str}/{solver_choice}/halo_{halo_size}"
os.makedirs(trace_folder, exist_ok=True)
run_simulation(mesh_shape, box_size, halo_size, solver_choice, iterations, hlo_print, trace, pdims, trace_folder)
print(f"Profiling done! Check {trace_folder}")
elif hlo_print:
hlo_folder = f"{output_path}/hlo/jaxpm/{nb_gpus}/{mesh_shape[0]}_{int(box_size[0])}/{pdm_str}/{solver_choice}/halo_{halo_size}"
os.makedirs(hlo_folder, exist_ok=True)
jaxpr , compiled , compiled2 = run_simulation(mesh_shape, box_size, halo_size, solver_choice, iterations, hlo_print, trace, pdims, hlo_folder)
print(f"type of memory analysis {type(compiled.memory_analysis())}")
print(f"memory analysis {compiled.memory_analysis()}")
print(f"memory analysis again {compiled2.memory_analysis()}")
jax.tree.map(lambda x: print(x), compiled.memory_analysis())
with open(f'{hlo_folder}/hlo_jaxpm.md', 'w') as f:
f.write(f"# JAXPM HLO\n")
f.write(f"## Args: {args}\n")
f.write(f"## JAXPR is \n")
f.write(f'---\n')
f.write(f"{jaxpr}\n")
f.write(f'---\n')
f.write(f"Lowered as text is \n")
f.write(f'---\n')
# f.write(f"{lower_as_text}\n")
f.write(f'---\n')
f.write(f"Compiled is \n")
f.write(f'---\n')
f.write(f"{compiled.as_text()}\n")
f.write(f"Cost analysis is \n")
f.write(f'---\n')
f.write(f"{compiled.cost_analysis()[0]['flops']}\n")
f.write(f'---\n')
f.write(f"Memory analysis is \n")
f.write(f'---\n')
f.write(f"{compiled.memory_analysis()}\n")
f.write(f'---\n')
print(f"Saved HLO to {hlo_folder}")
else:
final_field, stats, chrono_fun = run_simulation(mesh_shape, box_size, halo_size, solver_choice, iterations, hlo_print, trace, pdims, output_path)
print(f"shape of final_field {final_field.shape} and sharding spec {final_field.sharding} and local shape {final_field.addressable_data(0).shape}")
metadata = {
>>>>>>> glab/ASKabalan/jaxdecomp_proto
'rank': rank, 'rank': rank,
'function_name': f'JAXPM-{solver_choice}', 'function_name': f'JAXPM-{solver_choice}',
'precision': args.precision, 'precision': args.precision,
@ -236,6 +337,7 @@ if __name__ == "__main__":
'py': str(pdims[1]), 'py': str(pdims[1]),
'backend': 'NCCL', 'backend': 'NCCL',
'nodes': str(args.nodes) 'nodes': str(args.nodes)
<<<<<<< HEAD
} }
# Print the results to a CSV file # Print the results to a CSV file
chrono_fun.print_to_csv(f'{output_path}/jaxpm_benchmark.csv', **metadata) chrono_fun.print_to_csv(f'{output_path}/jaxpm_benchmark.csv', **metadata)
@ -254,8 +356,27 @@ if __name__ == "__main__":
if args.save_fields: if args.save_fields:
np.save(f'{field_folder}/final_field_0_{rank}.npy', np.save(f'{field_folder}/final_field_0_{rank}.npy',
final_field.addressable_data(0)) final_field.addressable_data(0))
=======
}
# Print the results to a CSV file
chrono_fun.print_to_csv(f'{output_path}/jaxpm_benchmark.csv', **metadata)
print(f"Finished! ") # Save the final field
print(f"Stats {stats}") >>>>>>> glab/ASKabalan/jaxdecomp_proto
print(f"Saving to {output_path}/jax_pm_benchmark.csv")
print(f"Saving field and logs in {field_folder}") field_folder = f"{output_path}/final_field/jaxpm/{nb_gpus}/{mesh_size}_{int(box_size[0])}/{pdm_str}/{solver_choice}/halo_{halo_size}"
os.makedirs(field_folder, exist_ok=True)
with open(f'{field_folder}/jaxpm.log', 'w') as f:
f.write(f"Args: {args}\n")
f.write(f"JIT time: {chrono_fun.jit_time:.4f} ms\n")
for i , time in enumerate(chrono_fun.times):
f.write(f"Time {i}: {time:.4f} ms\n")
f.write(f"Stats: {stats}\n")
if args.save_fields:
np.save(f'{field_folder}/final_field_0_{rank}.npy',
final_field.addressable_data(0))
print(f"Finished! ")
print(f"Stats {stats}")
print(f"Saving to {output_path}/jax_pm_benchmark.csv")
print(f"Saving field and logs in {field_folder}")

62
benchmarks/particle_mesh_a100.slurm → benchmarks/particle_mesh.slurm
View file

@ -1,40 +1,15 @@
#!/bin/bash #!/bin/bash
########################################## ##############################################################################################################################
## SELECT EITHER tkc@a100 OR tkc@v100 ## # USAGE:sbatch --account=tkc@a100 --nodes=1 --gres=gpu:1 --tasks-per-node=1 -C a100 benchmarks/particle_mesh_a100.slurm
########################################## ##############################################################################################################################
#SBATCH --account tkc@a100 #SBATCH --job-name=Particle-Mesh # nom du job
##########################################
#SBATCH --job-name=1N-FFT-Mesh # nom du job
# Il est possible d'utiliser une autre partition que celle par default
# en activant l'une des 5 directives suivantes :
##########################################
## SELECT EITHER a100 or v100-32g ##
##########################################
#SBATCH -C a100
##########################################
#******************************************
##########################################
## SELECT Number of nodes and GPUs per node
## For A100 ntasks-per-node and gres=gpu should be 8
## For V100 ntasks-per-node and gres=gpu should be 4
##########################################
#SBATCH --nodes=1 # nombre de noeud
#SBATCH --ntasks-per-node=8 # nombre de tache MPI par noeud (= nombre de GPU par noeud)
#SBATCH --gres=gpu:8 # nombre de GPU par nœud (max 8 avec gpu_p2, gpu_p5)
##########################################
## Le nombre de CPU par tache doit etre adapte en fonction de la partition utilisee. Sachant
## qu'ici on ne reserve qu'un seul GPU par tache (soit 1/4 ou 1/8 des GPU du noeud suivant
## la partition), l'ideal est de reserver 1/4 ou 1/8 des CPU du noeud pour chaque tache:
##########################################
#SBATCH --cpus-per-task=8 # nombre de CPU par tache pour gpu_p5 (1/8 du noeud 8-GPU) #SBATCH --cpus-per-task=8 # nombre de CPU par tache pour gpu_p5 (1/8 du noeud 8-GPU)
##########################################
# /!\ Attention, "multithread" fait reference a l'hyperthreading dans la terminologie Slurm
#SBATCH --hint=nomultithread # hyperthreading desactive #SBATCH --hint=nomultithread # hyperthreading desactive
#SBATCH --time=04:00:00 # temps d'execution maximum demande (HH:MM:SS) #SBATCH --time=04:00:00 # temps d'execution maximum demande (HH:MM:SS)
#SBATCH --output=%x_%N_a100.out # nom du fichier de sortie #SBATCH --output=%x_%N_a100.out # nom du fichier de sortie
#SBATCH --error=%x_%N_a100.out # nom du fichier d'erreur (ici commun avec la sortie) #SBATCH --error=%x_%N_a100.err # nom du fichier d'erreur (ici commun avec la sortie)
#SBATCH --exclusive # ressources dediees
##SBATCH --qos=qos_gpu-dev ##SBATCH --qos=qos_gpu-dev
## SBATCH --exclusive # ressources dediees
# Nettoyage des modules charges en interactif et herites par defaut # Nettoyage des modules charges en interactif et herites par defaut
num_nodes=$SLURM_JOB_NUM_NODES num_nodes=$SLURM_JOB_NUM_NODES
num_gpu_per_node=$SLURM_NTASKS_PER_NODE num_gpu_per_node=$SLURM_NTASKS_PER_NODE
@ -44,12 +19,11 @@ nb_gpus=$(( num_nodes * num_gpu_per_node))
module purge module purge
echo "Job constraint: $SLURM_JOB_CONSTRAINT"
echo "Job partition: $SLURM_JOB_PARTITION" echo "Job partition: $SLURM_JOB_PARTITION"
# Decommenter la commande module suivante si vous utilisez la partition "gpu_p5" # Decommenter la commande module suivante si vous utilisez la partition "gpu_p5"
# pour avoir acces aux modules compatibles avec cette partition # pour avoir acces aux modules compatibles avec cette partition
if [ $SLURM_JOB_PARTITION -eq gpu_p5 ]; then if [[ "$SLURM_JOB_PARTITION" == "gpu_p5" ]]; then
module load cpuarch/amd module load cpuarch/amd
source /gpfsdswork/projects/rech/tkc/commun/venv/a100/bin/activate source /gpfsdswork/projects/rech/tkc/commun/venv/a100/bin/activate
gpu_name=a100 gpu_name=a100
@ -66,8 +40,10 @@ module load nvidia-nsight-systems/2024.1.1.59
echo "The number of nodes allocated for this job is: $num_nodes" echo "The number of nodes allocated for this job is: $num_nodes"
echo "The number of GPUs allocated for this job is: $nb_gpus" echo "The number of GPUs allocated for this job is: $nb_gpus"
export EQX_ON_ERROR=nan
export ENABLE_PERFO_STEP=NVTX export ENABLE_PERFO_STEP=NVTX
export MPI4JAX_USE_CUDA_MPI=1 export MPI4JAX_USE_CUDA_MPI=1
function profile_python() { function profile_python() {
if [ $# -lt 1 ]; then if [ $# -lt 1 ]; then
echo "Usage: profile_python <python_script> [arguments for the script]" echo "Usage: profile_python <python_script> [arguments for the script]"
@ -75,14 +51,14 @@ function profile_python() {
fi fi
local script_name=$(basename "$1" .py) local script_name=$(basename "$1" .py)
local output_dir="prof_traces/$script_name" local output_dir="prof_traces/$gpu_name/$nb_gpus/$script_name"
local report_dir="out_prof/$gpu_name/$nb_gpus/$script_name" local report_dir="out_prof/$gpu_name/$nb_gpus/$script_name"
if [ $OUTPUT_FOLDER_ARGS -eq 1 ]; then if [ $OUTPUT_FOLDER_ARGS -eq 1 ]; then
local args=$(echo "${@:2}" | tr ' ' '_') local args=$(echo "${@:2}" | tr ' ' '_')
# Remove characters '/' and '-' from folder name # Remove characters '/' and '-' from folder name
args=$(echo "$args" | tr -d '/-') args=$(echo "$args" | tr -d '/-')
output_dir="prof_traces/$script_name/$args" output_dir="prof_traces/$gpu_name/$nb_gpus/$script_name/$args"
report_dir="out_prof/$gpu_name/$nb_gpus/$script_name/$args" report_dir="out_prof/$gpu_name/$nb_gpus/$script_name/$args"
fi fi
@ -99,13 +75,13 @@ function run_python() {
fi fi
local script_name=$(basename "$1" .py) local script_name=$(basename "$1" .py)
local output_dir="traces/$script_name" local output_dir="traces/$gpu_name/$nb_gpus/$script_name"
if [ $OUTPUT_FOLDER_ARGS -eq 1 ]; then if [ $OUTPUT_FOLDER_ARGS -eq 1 ]; then
local args=$(echo "${@:2}" | tr ' ' '_') local args=$(echo "${@:2}" | tr ' ' '_')
# Remove characters '/' and '-' from folder name # Remove characters '/' and '-' from folder name
args=$(echo "$args" | tr -d '/-') args=$(echo "$args" | tr -d '/-')
output_dir="traces/$script_name/$args" output_dir="traces/$gpu_name/$nb_gpus/$script_name/$args"
fi fi
mkdir -p "$output_dir" mkdir -p "$output_dir"
@ -142,12 +118,11 @@ pdims_table[8]="2x4 1x8 8x1 4x2"
pdims_table[16]="4x4 1x16 16x1" pdims_table[16]="4x4 1x16 16x1"
pdims_table[32]="4x8 8x4 1x32 32x1" pdims_table[32]="4x8 8x4 1x32 32x1"
pdims_table[64]="8x8 16x4 1x64 64x1" pdims_table[64]="8x8 16x4 1x64 64x1"
pdims_table[128]="8x16 16x8 4x32 32x4 1x128 128x1 2x64 64x2" pdims_table[128]="8x16 16x8 1x128 128x1"
pdims_table[160]="8x20 20x8 16x10 10x16 5x32 32x5 1x160 160x1 2x80 80x2 4x40 40x4" pdims_table[256]="16x16 1x256 256x1"
# mpch=(128 256 512 1024 2048 4096) # mpch=(128 256 512 1024 2048 4096)
grid=(256 512 1024 2048 4096) grid=(256 512 1024 2048 4096 8192)
precisions=(float32 float64) precisions=(float32 float64)
pdim="${pdims_table[$nb_gpus]}" pdim="${pdims_table[$nb_gpus]}"
solvers=(lpt lfm) solvers=(lpt lfm)
@ -164,8 +139,9 @@ fi
# GPU name is a100 if num_gpu_per_node is 8, otherwise it is v100 # GPU name is a100 if num_gpu_per_node is 8, otherwise it is v100
out_dir="pm_prof/$gpu_name/$nb_gpus" out_dir="pm_prof/$gpu_name/$nb_gpus"
trace_dir="traces/$gpu_name/$nb_gpus/bench_pm"
echo "Output dir is : $out_dir" echo "Output dir is : $out_dir"
echo "Trace dir is : $trace_dir"
for pr in "${precisions[@]}"; do for pr in "${precisions[@]}"; do
for g in "${grid[@]}"; do for g in "${grid[@]}"; do
@ -175,5 +151,7 @@ for pr in "${precisions[@]}"; do
slaunch bench_pm.py -m $g -b $g -p $p -hs $halo_size -pr $pr -s $solver -i 4 -o $out_dir -f -n $num_nodes slaunch bench_pm.py -m $g -b $g -p $p -hs $halo_size -pr $pr -s $solver -i 4 -o $out_dir -f -n $num_nodes
done done
done done
# delete crash core dump files
rm -f core.python.*
done done
done done

70
benchmarks/pmwd_v100.slurm → benchmarks/pmwd_pm.slurm
View file

@ -1,40 +1,15 @@
#!/bin/bash #!/bin/bash
########################################## ##############################################################################################################################
## SELECT EITHER tkc@a100 OR tkc@v100 ## # USAGE:sbatch --account=tkc@a100 --nodes=1 --gres=gpu:1 --tasks-per-node=1 -C a100 benchmarks/particle_mesh_a100.slurm
########################################## ##############################################################################################################################
#SBATCH --account tkc@v100 #SBATCH --job-name=Particle-Mesh # nom du job
##########################################
#SBATCH --job-name=16N-V100Particle-Mesh # nom du job
# Il est possible d'utiliser une autre partition que celle par default
# en activant l'une des 5 directives suivantes :
##########################################
## SELECT EITHER a100 or v100-32g ##
##########################################
#SBATCH -C v100-32g
##########################################
#******************************************
##########################################
## SELECT Number of nodes and GPUs per node
## For A100 ntasks-per-node and gres=gpu should be 8
## For V100 ntasks-per-node and gres=gpu should be 4
##########################################
#SBATCH --nodes=1 # nombre de noeud
#SBATCH --ntasks-per-node=1 # nombre de tache MPI par noeud (= nombre de GPU par noeud)
#SBATCH --gres=gpu:1 # nombre de GPU par nœud (max 8 avec gpu_p2, gpu_p5)
##########################################
## Le nombre de CPU par tache doit etre adapte en fonction de la partition utilisee. Sachant
## qu'ici on ne reserve qu'un seul GPU par tache (soit 1/4 ou 1/8 des GPU du noeud suivant
## la partition), l'ideal est de reserver 1/4 ou 1/8 des CPU du noeud pour chaque tache:
##########################################
#SBATCH --cpus-per-task=8 # nombre de CPU par tache pour gpu_p5 (1/8 du noeud 8-GPU) #SBATCH --cpus-per-task=8 # nombre de CPU par tache pour gpu_p5 (1/8 du noeud 8-GPU)
##########################################
# /!\ Attention, "multithread" fait reference a l'hyperthreading dans la terminologie Slurm
#SBATCH --hint=nomultithread # hyperthreading desactive #SBATCH --hint=nomultithread # hyperthreading desactive
#SBATCH --time=02:00:00 # temps d'execution maximum demande (HH:MM:SS) #SBATCH --time=04:00:00 # temps d'execution maximum demande (HH:MM:SS)
#SBATCH --output=%x_%N_a100.out # nom du fichier de sortie #SBATCH --output=%x_%N_a100.out # nom du fichier de sortie
#SBATCH --error=%x_%N_a100.out # nom du fichier d'erreur (ici commun avec la sortie) #SBATCH --error=%x_%N_a100.out # nom du fichier d'erreur (ici commun avec la sortie)
#SBATCH --qos=qos_gpu-dev
#SBATCH --exclusive # ressources dediees #SBATCH --exclusive # ressources dediees
##SBATCH --qos=qos_gpu-dev
# Nettoyage des modules charges en interactif et herites par defaut # Nettoyage des modules charges en interactif et herites par defaut
num_nodes=$SLURM_JOB_NUM_NODES num_nodes=$SLURM_JOB_NUM_NODES
num_gpu_per_node=$SLURM_NTASKS_PER_NODE num_gpu_per_node=$SLURM_NTASKS_PER_NODE
@ -44,12 +19,11 @@ nb_gpus=$(( num_nodes * num_gpu_per_node))
module purge module purge
echo "Job constraint: $SLURM_JOB_CONSTRAINT"
echo "Job partition: $SLURM_JOB_PARTITION" echo "Job partition: $SLURM_JOB_PARTITION"
# Decommenter la commande module suivante si vous utilisez la partition "gpu_p5" # Decommenter la commande module suivante si vous utilisez la partition "gpu_p5"
# pour avoir acces aux modules compatibles avec cette partition # pour avoir acces aux modules compatibles avec cette partition
if [ $SLURM_JOB_PARTITION -eq gpu_p5 ]; then if [[ "$SLURM_JOB_PARTITION" == "gpu_p5" ]]; then
module load cpuarch/amd module load cpuarch/amd
source /gpfsdswork/projects/rech/tkc/commun/venv/a100/bin/activate source /gpfsdswork/projects/rech/tkc/commun/venv/a100/bin/activate
gpu_name=a100 gpu_name=a100
@ -67,9 +41,9 @@ echo "The number of nodes allocated for this job is: $num_nodes"
echo "The number of GPUs allocated for this job is: $nb_gpus" echo "The number of GPUs allocated for this job is: $nb_gpus"
export EQX_ON_ERROR=nan export EQX_ON_ERROR=nan
export CUDA_ALLOC=1
export ENABLE_PERFO_STEP=NVTX export ENABLE_PERFO_STEP=NVTX
export MPI4JAX_USE_CUDA_MPI=1 export MPI4JAX_USE_CUDA_MPI=1
function profile_python() { function profile_python() {
if [ $# -lt 1 ]; then if [ $# -lt 1 ]; then
echo "Usage: profile_python <python_script> [arguments for the script]" echo "Usage: profile_python <python_script> [arguments for the script]"
@ -77,14 +51,14 @@ function profile_python() {
fi fi
local script_name=$(basename "$1" .py) local script_name=$(basename "$1" .py)
local output_dir="prof_traces/$script_name" local output_dir="prof_traces/$gpu_name/$nb_gpus/$script_name"
local report_dir="out_prof/$gpu_name/$nb_gpus/$script_name" local report_dir="out_prof/$gpu_name/$nb_gpus/$script_name"
if [ $OUTPUT_FOLDER_ARGS -eq 1 ]; then if [ $OUTPUT_FOLDER_ARGS -eq 1 ]; then
local args=$(echo "${@:2}" | tr ' ' '_') local args=$(echo "${@:2}" | tr ' ' '_')
# Remove characters '/' and '-' from folder name # Remove characters '/' and '-' from folder name
args=$(echo "$args" | tr -d '/-') args=$(echo "$args" | tr -d '/-')
output_dir="prof_traces/$script_name/$args" output_dir="prof_traces/$gpu_name/$nb_gpus/$script_name/$args"
report_dir="out_prof/$gpu_name/$nb_gpus/$script_name/$args" report_dir="out_prof/$gpu_name/$nb_gpus/$script_name/$args"
fi fi
@ -101,13 +75,13 @@ function run_python() {
fi fi
local script_name=$(basename "$1" .py) local script_name=$(basename "$1" .py)
local output_dir="traces/$script_name" local output_dir="traces/$gpu_name/$nb_gpus/$script_name"
if [ $OUTPUT_FOLDER_ARGS -eq 1 ]; then if [ $OUTPUT_FOLDER_ARGS -eq 1 ]; then
local args=$(echo "${@:2}" | tr ' ' '_') local args=$(echo "${@:2}" | tr ' ' '_')
# Remove characters '/' and '-' from folder name # Remove characters '/' and '-' from folder name
args=$(echo "$args" | tr -d '/-') args=$(echo "$args" | tr -d '/-')
output_dir="traces/$script_name/$args" output_dir="traces/$gpu_name/$nb_gpus/$script_name/$args"
fi fi
mkdir -p "$output_dir" mkdir -p "$output_dir"
@ -116,6 +90,7 @@ function run_python() {
} }
# run or profile # run or profile
function slaunch() { function slaunch() {
@ -136,13 +111,9 @@ export TMPDIR=$JOBSCRATCH
# faire pointer le répertoire /tmp/nvidia vers TMPDIR # faire pointer le répertoire /tmp/nvidia vers TMPDIR
ln -s $JOBSCRATCH /tmp/nvidia ln -s $JOBSCRATCH /tmp/nvidia
# mpch=(128 256 512 1024 2048 4096) # mpch=(128 256 512 1024 2048 4096)
grid=(256 512 1024 2048 4096) grid=(256 512 1024 2048 4096 8192)
precisions=(float32 float64) precisions=(float32 float64)
pdim="${pdims_table[$nb_gpus]}"
solvers=(lpt lfm) solvers=(lpt lfm)
# GPU name is a100 if num_gpu_per_node is 8, otherwise it is v100 # GPU name is a100 if num_gpu_per_node is 8, otherwise it is v100
@ -154,14 +125,23 @@ else
fi fi
out_dir="pm_prof/$gpu_name/$nb_gpus" out_dir="pm_prof/$gpu_name/$nb_gpus"
trace_dir="traces/$gpu_name/$nb_gpus/bench_pmwd"
echo "Output dir is : $out_dir" echo "Output dir is : $out_dir"
echo "Trace dir is : $trace_dir"
for pr in "${precisions[@]}"; do for pr in "${precisions[@]}"; do
for g in "${grid[@]}"; do for g in "${grid[@]}"; do
for solver in "${solvers[@]}"; do for solver in "${solvers[@]}"; do
slaunch bench_pmwd.py -m $g -b $g -pr $pr -s $solver -i 4 -o $out_dir -f slaunch bench_pmwd.py -m $g -b $g -pr $pr -s $solver -i 4 -o $out_dir -f
done done
# delete crash core dump files
rm -f core.python.*
done done
done done
# # zip the output files and traces
# tar -czvf $out_dir.tar.gz $out_dir
# tar -czvf $trace_dir.tar.gz $trace_dir
# # remove the output files and traces
# rm -rf $out_dir $trace_dir
#

18
benchmarks/run_all_jobs.sh
View file

@ -1,19 +1,21 @@
#!/bin/bash #!/bin/bash
# Run all slurms jobs # Run all slurms jobs
nodes_v100=(1 2 4 8 16) nodes_v100=(1 2 4 8 16 32)
nodes_a100=(1 2 4 8 16) nodes_a100=(1 2 4 8 16 32)
for n in ${nodes_v100[@]}; do for n in ${nodes_v100[@]}; do
sbatch --nodes=$n --job-name=v100_$n-JAXPM particle_mesh_v100.slurm sbatch --account=tkc@v100 --nodes=$n --gres=gpu:4 --tasks-per-node=4 -C v100-32g --job-name=JAXPM-$n-N-v100 particle_mesh.slurm
done done
for n in ${nodes_a100[@]}; do for n in ${nodes_a100[@]}; do
sbatch --nodes=$n --job-name=a100_$n-JAXPM particle_mesh_a100.slurm sbatch --account=tkc@a100 --nodes=$n --gres=gpu:4 --tasks-per-node=4 -C a100 --job-name=JAXPM-$n-N-a100 particle_mesh.slurm
done done
# single GPUs # single GPUs
sbatch --job-name=JAXPM-1GPU-V100 --nodes=1 --gres=gpu:1 --tasks-per-node=1 particle_mesh_v100.slurm sbatch --account=tkc@a100 --nodes=1 --gres=gpu:1 --tasks-per-node=1 -C a100 --job-name=JAXPM-1GPU-V100 particle_mesh.slurm
sbatch --job-name=JAXPM-1GPU-A100 --nodes=1 --gres=gpu:1 --tasks-per-node=1 particle_mesh_a100.slurm sbatch --account=tkc@v100 --nodes=1 --gres=gpu:1 --tasks-per-node=1 -C v100-32g --job-name=JAXPM-1GPU-A100 particle_mesh.slurm
sbatch --job-name=PMWD-v100 pmwd_v100.slurm sbatch --account=tkc@a100 --nodes=1 --gres=gpu:1 --tasks-per-node=1 -C a100 --job-name=PMWD-1GPU-v100 pmwd_pm.slurm
sbatch --job-name=PMWD-a100 pmwd_a100.slurm sbatch --account=tkc@v100 --nodes=1 --gres=gpu:1 --tasks-per-node=1 -C v100-32g --job-name=PMWD-1GPU-a100 pmwd_pm.slurm

4
jaxpm/distributed.py
View file

@ -91,8 +91,8 @@ def get_halo_size(halo_size):
def halo_exchange(x, halo_extents, halo_periods=(True, True, True)): def halo_exchange(x, halo_extents, halo_periods=(True, True, True)):
mesh = mesh_lib.thread_resources.env.physical_mesh mesh = mesh_lib.thread_resources.env.physical_mesh
if distributed and not (mesh.empty) and (halo_extents[0] > 0 if distributed and not (mesh.empty) and (halo_extents > 0
or halo_extents[1] > 0): or halo_extents > 0):
return jaxdecomp.halo_exchange(x, halo_extents, halo_periods) return jaxdecomp.halo_exchange(x, halo_extents, halo_periods)
else: else:
return x return x

41
jaxpm/painting.py
View file

@ -50,15 +50,15 @@ def cic_paint_impl(mesh, displacement, weight=None):
@partial(jax.jit, static_argnums=(2, )) @partial(jax.jit, static_argnums=(2, ))
def cic_paint(mesh, positions, halo_size=0, weight=None): def cic_paint(mesh, positions, halo_size=0, weight=None):
halo_size, halo_extents = get_halo_size(halo_size) halo_padding, halo_extents = get_halo_size(halo_size)
mesh = slice_pad(mesh, halo_size) mesh = slice_pad(mesh, halo_padding)
mesh = autoshmap(cic_paint_impl, mesh = autoshmap(cic_paint_impl,
in_specs=(P('x', 'y'), P('x', 'y'), P()), in_specs=(P('x', 'y'), P('x', 'y'), P()),
out_specs=P('x', 'y'))(mesh, positions, weight) out_specs=P('x', 'y'))(mesh, positions, weight)
mesh = halo_exchange(mesh, mesh = halo_exchange(mesh,
halo_extents=halo_extents, halo_extents=halo_size // 2,
halo_periods=(True, True, True)) halo_periods=True)
mesh = slice_unpad(mesh, halo_size) mesh = slice_unpad(mesh, halo_padding)
return mesh return mesh
@ -95,11 +95,11 @@ def cic_read_impl(mesh, displacement):
@partial(jax.jit, static_argnums=(2, )) @partial(jax.jit, static_argnums=(2, ))
def cic_read(mesh, displacement, halo_size=0): def cic_read(mesh, displacement, halo_size=0):
halo_size, halo_extents = get_halo_size(halo_size) halo_padding, halo_extents = get_halo_size(halo_size)
mesh = slice_pad(mesh, halo_size) mesh = slice_pad(mesh, halo_padding)
mesh = halo_exchange(mesh, mesh = halo_exchange(mesh,
halo_extents=halo_extents, halo_extents=halo_size//2,
halo_periods=(True, True, True)) halo_periods=True)
displacement = autoshmap(cic_read_impl, displacement = autoshmap(cic_read_impl,
in_specs=(P('x', 'y'), P('x', 'y')), in_specs=(P('x', 'y'), P('x', 'y')),
out_specs=P('x', 'y'))(mesh, displacement) out_specs=P('x', 'y'))(mesh, displacement)
@ -159,17 +159,24 @@ def cic_paint_dx_impl(displacements, halo_size):
@partial(jax.jit, static_argnums=(1, )) @partial(jax.jit, static_argnums=(1, ))
def cic_paint_dx(displacements, halo_size=0): def cic_paint_dx(displacements, halo_size=0):
<<<<<<< HEAD
halo_size, halo_extents = get_halo_size(halo_size) halo_size, halo_extents = get_halo_size(halo_size)
mesh = autoshmap(partial(cic_paint_dx_impl, halo_size=halo_size), mesh = autoshmap(partial(cic_paint_dx_impl, halo_size=halo_size),
=======
halo_padding, halo_extents = get_halo_size(halo_size)
mesh = autoshmap(partial(cic_paint_dx_impl, halo_size=halo_padding),
>>>>>>> glab/ASKabalan/jaxdecomp_proto
in_specs=(P('x', 'y')), in_specs=(P('x', 'y')),
out_specs=P('x', 'y'))(displacements) out_specs=P('x', 'y'))(displacements)
mesh = halo_exchange(mesh, mesh = halo_exchange(mesh,
halo_extents=halo_extents, halo_extents=halo_size//2,
halo_periods=(True, True, True)) halo_periods=True)
mesh = slice_unpad(mesh, halo_size) mesh = slice_unpad(mesh, halo_padding)
return mesh return mesh
@ -196,12 +203,18 @@ def cic_read_dx_impl(mesh, halo_size):
@partial(jax.jit, static_argnums=(1, )) @partial(jax.jit, static_argnums=(1, ))
def cic_read_dx(mesh, halo_size=0): def cic_read_dx(mesh, halo_size=0):
# return mesh # return mesh
halo_size, halo_extents = get_halo_size(halo_size) halo_padding, halo_extents = get_halo_size(halo_size)
mesh = slice_pad(mesh, halo_size) mesh = slice_pad(mesh, halo_padding)
mesh = halo_exchange(mesh, mesh = halo_exchange(mesh,
<<<<<<< HEAD
halo_extents=halo_extents, halo_extents=halo_extents,
halo_periods=(True, True, True)) halo_periods=(True, True, True))
displacements = autoshmap(partial(cic_read_dx_impl, halo_size=halo_size), displacements = autoshmap(partial(cic_read_dx_impl, halo_size=halo_size),
=======
halo_extents=halo_size//2,
halo_periods=True)
displacements = autoshmap(partial(cic_read_dx_impl , halo_size=halo_padding),
>>>>>>> glab/ASKabalan/jaxdecomp_proto
in_specs=(P('x', 'y')), in_specs=(P('x', 'y')),
out_specs=P('x', 'y'))(mesh) out_specs=P('x', 'y'))(mesh)

5
scripts/particle_mesh.slurm
View file

@ -62,8 +62,8 @@ module load nvidia-nsight-systems/2024.1.1.59
echo "The number of nodes allocated for this job is: $num_nodes" echo "The number of nodes allocated for this job is: $num_nodes"
echo "The number of GPUs allocated for this job is: $nb_gpus" echo "The number of GPUs allocated for this job is: $nb_gpus"
export EQX_ON_ERROR=nan export ENABLE_PERFO_STEP=NVTX
export CUDA_ALLOC=1 export MPI4JAX_USE_CUDA_MPI=1
function profile_python() { function profile_python() {
if [ $# -lt 1 ]; then if [ $# -lt 1 ]; then
@ -122,6 +122,7 @@ set -x
declare -A pdims_table declare -A pdims_table
# Define the table # Define the table
pdims_table[1]="1x1"
pdims_table[4]="2x2 1x4" pdims_table[4]="2x2 1x4"
pdims_table[8]="2x4 1x8" pdims_table[8]="2x4 1x8"
pdims_table[16]="2x8 1x16" pdims_table[16]="2x8 1x16"