ClimaAtmos - Target GPU simulations

Public

Monitor performance of target GPU simulations

Try latest main

#299
/
main/4de11009
Failed in 1h 37m
buildkite-agent pipeline upload .buildkite/gpu_...
init :GPU:
Charles Kawczynski
Charles Kawczynski
Created
Triggered from Web
dry baroclinic wavemkdir -p target_gpu_implicit_baroclinic_wave && nsys profile --trace=nvtx,mpi,cuda,osrt --output=target_gpu_implicit_baroclinic_wave/output_active/report julia --threads=3 --color=yes --project=examples examples/hybrid/driver.jl --config_file config/gpu_configs/target_gpu_implicit_baroclinic_wave.yml --job_id target_gpu_implicit_baroclinic_wave
Waited 13s
·
Ran in 1m 8s
Agent
moist Held-Suarezmkdir -p gpu_hs_rhoe_equil_55km_nz63_0M && nsys profile --trace=nvtx,mpi,cuda,osrt --output=gpu_hs_rhoe_equil_55km_nz63_0M/output_active/report julia --threads=3 --color=yes --project=examples examples/hybrid/driver.jl --config_file config/gpu_configs/gpu_hs_rhoe_equil_55km_nz63_0M.yml --job_id gpu_hs_rhoe_equil_55km_nz63_0M
Waited 13s
·
Ran in 1m 7s
Agent
moist Held-Suarez - 4 gpusmkdir -p gpu_hs_rhoe_equil_55km_nz63_0M_4process && srun --cpu-bind=threads --cpus-per-task=4 nsys profile --trace=nvtx,mpi,cuda,osrt --output=gpu_hs_rhoe_equil_55km_nz63_0M_4process/output_active/report-%q{PMI_RANK} julia --threads=3 --color=yes --project=examples examples/hybrid/driver.jl --config_file config/gpu_configs/gpu_hs_rhoe_equil_55km_nz63_0M_4process.yml --job_id gpu_hs_rhoe_equil_55km_nz63_0M_4process
Waited 13s
·
Ran in 1m 14s
Agent
dry baroclinic wave - 4 gpusmkdir -p target_gpu_implicit_baroclinic_wave_4process && srun --cpu-bind=threads --cpus-per-task=4 nsys profile --trace=osrt,nvtx,cuda,mpi,ucx --output=target_gpu_implicit_baroclinic_wave_4process/output_active/report-%q{PMI_RANK} julia --threads=3 --color=yes --project=examples examples/hybrid/driver.jl --config_file config/gpu_configs/target_gpu_implicit_baroclinic_wave_4process.yml --job_id target_gpu_implicit_baroclinic_wave_4process
Waited 1m 29s
·
Ran in 1m 4s
Agent
gpu_aquaplanet_dyamond - strong scaling - 1 GPUmkdir -p gpu_aquaplanet_dyamond_ss_1process && srun --cpu-bind=threads --cpus-per-task=4 nsys profile --trace=nvtx,mpi,cuda,osrt --output=gpu_aquaplanet_dyamond_ss_1process/output_active/report julia --threads=3 --color=yes --project=examples examples/hybrid/driver.jl --config_file config/gpu_configs/gpu_aquaplanet_dyamond_ss_1process.yml --job_id gpu_aquaplanet_dyamond_ss_1process
Waited 1m 29s
·
Ran in 1m 6s
Agent
gpu_aquaplanet_progedmf_benchmarkjulia --color=yes --project=perf perf/benchmark.jl --config_file config/model_configs/aquaplanet_progedmf.yml --job_id gpu_aquaplanet_progedmf_benchmark
Waited 20m 51s
·
Ran in 1h 8m
Agent
Total Job Run Time: 3h 17m