Strong scaling performance, Sphere benchmark, K80, 64M particles
Performance in millions of particle-timesteps / second / node

Nodes Kokkos/Cuda-1 (mpi) Kokkos/Cuda-2 (mpi)
1 159.4 (2) 334.6 (4)
2 161 (2) 317.3 (4)
4 128.2 (2) 273.3 (4)
8 None 231.1 (4)

Run commands and logfile links for column Kokkos/Cuda-1

1 mpirun -np 4 --npersocket 1 --bind-to core spa_ride80_kokkos_cuda -sf kk -k on g 2 -pk kokkos reduction atomic comm threaded -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.gpu.steps -log log.sparta.date=23Dec17.model=sphere.machine=ride80.pkg=kokkos_cuda.kind=strong.size=64M.node=2.mpi=2.gpu=2
2 mpirun -np 8 --npersocket 1 --bind-to core spa_ride80_kokkos_cuda -sf kk -k on g 2 -pk kokkos reduction atomic comm threaded -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.gpu.steps -log log.sparta.date=23Dec17.model=sphere.machine=ride80.pkg=kokkos_cuda.kind=strong.size=64M.node=4.mpi=2.gpu=2
4 mpirun -np 16 --npersocket 1 --bind-to core spa_ride80_kokkos_cuda -sf kk -k on g 2 -pk kokkos reduction atomic comm threaded -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.gpu.steps -log log.sparta.date=23Dec17.model=sphere.machine=ride80.pkg=kokkos_cuda.kind=strong.size=64M.node=8.mpi=2.gpu=2
8 None

Run commands and logfile links for column Kokkos/Cuda-2

1 mpirun -np 4 --npersocket 2 --bind-to core spa_ride80_kokkos_cuda -sf kk -k on g 4 -pk kokkos reduction atomic comm threaded -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.gpu.steps -log log.sparta.date=23Dec17.model=sphere.machine=ride80.pkg=kokkos_cuda.kind=strong.size=64M.node=1.mpi=4.gpu=4
2 mpirun -np 8 --npersocket 2 --bind-to core spa_ride80_kokkos_cuda -sf kk -k on g 4 -pk kokkos reduction atomic comm threaded -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.gpu.steps -log log.sparta.date=23Dec17.model=sphere.machine=ride80.pkg=kokkos_cuda.kind=strong.size=64M.node=2.mpi=4.gpu=4
4 mpirun -np 16 --npersocket 2 --bind-to core spa_ride80_kokkos_cuda -sf kk -k on g 4 -pk kokkos reduction atomic comm threaded -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.gpu.steps -log log.sparta.date=23Dec17.model=sphere.machine=ride80.pkg=kokkos_cuda.kind=strong.size=64M.node=4.mpi=4.gpu=4
8 mpirun -np 32 --npersocket 2 --bind-to core spa_ride80_kokkos_cuda -sf kk -k on g 4 -pk kokkos reduction atomic comm threaded -v x 256 -v y 160 -v z 160 -v t 100 -in in.sphere.gpu.steps -log log.sparta.date=23Dec17.model=sphere.machine=ride80.pkg=kokkos_cuda.kind=strong.size=64M.node=8.mpi=4.gpu=4