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

Nodes	CPU/KNL (mpi,hyper)	Kokkos/KNL (mpi,thread,hyper)	Kokkos/serial/KNL (mpi,hyper)
1	244.4 (256,4)	239.8 (32,8,4)	255.7 (256,4)
2	247.8 (256,4)	253.2 (128,2,4)	263.9 (256,4)
4	250.7 (256,4)	264 (64,4,4)	271.4 (256,4)
8	242.4 (256,4)	277.1 (128,2,4)	275.9 (256,4)
16	221.7 (128,2)	287.9 (64,4,4)	247.4 (256,4)
32	233.2 (256,4)	287.6 (64,4,4)	231.7 (256,4)
64	233.3 (128,2)	239.3 (32,8,4)	211.7 (128,2)

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 256 -v y 160 -v z 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=cpu_knl.kind=strong.size=64M.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 256 -v y 160 -v z 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=cpu_knl.kind=strong.size=64M.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.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=cpu_knl.kind=strong.size=64M.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 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=cpu_knl.kind=strong.size=64M.node=8.mpi=256.hyper=4
16	srun -n 2048 -C knl --ntasks-per-node 128 --cpu_bind=threads -c 2 ./spa_mutrino_knl -v x 256 -v y 160 -v z 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=cpu_knl.kind=strong.size=64M.node=16.mpi=128.hyper=2
32	srun -n 8192 -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.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=cpu_knl.kind=strong.size=64M.node=32.mpi=256.hyper=4
64	srun -n 8192 -C knl --ntasks-per-node 128 --cpu_bind=threads -c 2 ./spa_mutrino_knl -v x 256 -v y 160 -v z 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=cpu_knl.kind=strong.size=64M.node=64.mpi=128.hyper=2

Run commands and logfile links for column Kokkos/KNL

1	setenv OMP_NUM_THREADS 8; srun -n 32 -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 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_knl.kind=strong.size=64M.node=1.mpi=32.thread=8.hyper=4
2	setenv OMP_NUM_THREADS 2; srun -n 256 -C knl --ntasks-per-node 128 --cpu_bind=threads -c 2 ./spa_mutrino_kokkos_knl -sf kk -k on t 2 -pk kokkos reduction parallel/reduce comm classic -v x 256 -v y 160 -v z 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_knl.kind=strong.size=64M.node=2.mpi=128.thread=2.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.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_knl.kind=strong.size=64M.node=4.mpi=64.thread=4.hyper=4
8	setenv OMP_NUM_THREADS 2; srun -n 1024 -C knl --ntasks-per-node 128 --cpu_bind=threads -c 2 ./spa_mutrino_kokkos_knl -sf kk -k on t 2 -pk kokkos reduction parallel/reduce comm classic -v x 256 -v y 160 -v z 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_knl.kind=strong.size=64M.node=8.mpi=128.thread=2.hyper=4
16	setenv OMP_NUM_THREADS 4; srun -n 1024 -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.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_knl.kind=strong.size=64M.node=16.mpi=64.thread=4.hyper=4
32	setenv OMP_NUM_THREADS 4; srun -n 2048 -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.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_knl.kind=strong.size=64M.node=32.mpi=64.thread=4.hyper=4
64	setenv OMP_NUM_THREADS 8; srun -n 2048 -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 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_knl.kind=strong.size=64M.node=64.mpi=32.thread=8.hyper=4

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 256 -v y 160 -v z 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_serial_knl.kind=strong.size=64M.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 256 -v y 160 -v z 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_serial_knl.kind=strong.size=64M.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.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_serial_knl.kind=strong.size=64M.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 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_serial_knl.kind=strong.size=64M.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 160 -v z 160 -v t 100 -in in.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_serial_knl.kind=strong.size=64M.node=16.mpi=256.hyper=4
32	srun -n 8192 -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.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_serial_knl.kind=strong.size=64M.node=32.mpi=256.hyper=4
64	srun -n 8192 -C knl --ntasks-per-node 128 --cpu_bind=threads -c 2 ./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.collide.steps -log log.sparta.date=23Dec17.model=collide.machine=mutrino.pkg=kokkos_serial_knl.kind=strong.size=64M.node=64.mpi=128.hyper=2