Weak scaling performance, Sphere benchmark, KNL, 16M particles/node
Performance in millions of particle-timesteps / second / node

Nodes CPU/KNL (mpi,hyper) Kokkos/KNL (mpi,thread,hyper) Kokkos/serial/KNL (mpi,hyper)
1 252.3 (256,4) 268.6 (64,4,4) 252.3 (256,4)
2 226.2 (256,4) 265.6 (64,4,4) 225.5 (256,4)
4 217.6 (256,4) 222.2 (64,4,4) 217.7 (256,4)
8 214.2 (256,4) 208.2 (32,8,4) 193.9 (256,4)
16 199.6 (256,4) 142.8 (64,2,2) 197.4 (256,4)

Run commands and logfile links for column CPU/KNL

1 srun -n 256 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_knl -v x 128 -v y 80 -v z 160 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=cpu_knl.kind=weak.size=16M.node=1.mpi=256.hyper=4
2 srun -n 512 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_knl -v x 128 -v y 160 -v z 160 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=cpu_knl.kind=weak.size=16M.node=2.mpi=256.hyper=4
4 srun -n 1024 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_knl -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=cpu_knl.kind=weak.size=16M.node=4.mpi=256.hyper=4
8 srun -n 2048 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_knl -v x 256 -v y 160 -v z 320 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=cpu_knl.kind=weak.size=16M.node=8.mpi=256.hyper=4
16 srun -n 4096 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_knl -v x 256 -v y 320 -v z 320 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=cpu_knl.kind=weak.size=16M.node=16.mpi=256.hyper=4

Run commands and logfile links for column Kokkos/KNL

1 setenv OMP_NUM_THREADS 4; srun -n 64 -C knl --ntasks-per-node 64 --cpu_bind=cores -c 4 ./spa_mutrino_kokkos_knl -sf kk -k on t 4 -pk kokkos reduction parallel/reduce comm classic -v x 128 -v y 80 -v z 160 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_knl.kind=weak.size=16M.node=1.mpi=64.thread=4.hyper=4
2 setenv OMP_NUM_THREADS 4; srun -n 128 -C knl --ntasks-per-node 64 --cpu_bind=cores -c 4 ./spa_mutrino_kokkos_knl -sf kk -k on t 4 -pk kokkos reduction parallel/reduce comm classic -v x 128 -v y 160 -v z 160 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_knl.kind=weak.size=16M.node=2.mpi=64.thread=4.hyper=4
4 setenv OMP_NUM_THREADS 4; srun -n 256 -C knl --ntasks-per-node 64 --cpu_bind=cores -c 4 ./spa_mutrino_kokkos_knl -sf kk -k on t 4 -pk kokkos reduction parallel/reduce comm classic -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_knl.kind=weak.size=16M.node=4.mpi=64.thread=4.hyper=4
8 setenv OMP_NUM_THREADS 8; srun -n 256 -C knl --ntasks-per-node 32 --cpu_bind=cores -c 8 ./spa_mutrino_kokkos_knl -sf kk -k on t 8 -pk kokkos reduction parallel/reduce comm classic -v x 256 -v y 160 -v z 320 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_knl.kind=weak.size=16M.node=8.mpi=32.thread=8.hyper=4
16 setenv OMP_NUM_THREADS 2; srun -n 1024 -C knl --ntasks-per-node 64 --cpu_bind=cores -c 4 ./spa_mutrino_kokkos_knl -sf kk -k on t 2 -pk kokkos reduction parallel/reduce comm classic -v x 256 -v y 320 -v z 320 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_knl.kind=weak.size=16M.node=16.mpi=64.thread=2.hyper=2

Run commands and logfile links for column Kokkos/serial/KNL

1 srun -n 256 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_kokkos_serial_knl -sf kk -k on -pk kokkos reduction parallel/reduce comm classic -v x 128 -v y 80 -v z 160 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_serial_knl.kind=weak.size=16M.node=1.mpi=256.hyper=4
2 srun -n 512 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_kokkos_serial_knl -sf kk -k on -pk kokkos reduction parallel/reduce comm classic -v x 128 -v y 160 -v z 160 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_serial_knl.kind=weak.size=16M.node=2.mpi=256.hyper=4
4 srun -n 1024 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_kokkos_serial_knl -sf kk -k on -pk kokkos reduction parallel/reduce comm classic -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_serial_knl.kind=weak.size=16M.node=4.mpi=256.hyper=4
8 srun -n 2048 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_kokkos_serial_knl -sf kk -k on -pk kokkos reduction parallel/reduce comm classic -v x 256 -v y 160 -v z 320 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_serial_knl.kind=weak.size=16M.node=8.mpi=256.hyper=4
16 srun -n 4096 -C knl --ntasks-per-node 256 --cpu_bind=threads -c 1 ./spa_mutrino_kokkos_serial_knl -sf kk -k on -pk kokkos reduction parallel/reduce comm classic -v x 256 -v y 320 -v z 320 -v t 100 -in in.sphere.steps -log log.sparta.date=23Dec17.model=sphere.machine=mutrino.pkg=kokkos_serial_knl.kind=weak.size=16M.node=16.mpi=256.hyper=4