[[wg:dynamo:Performance_results|Back to performance benchmark lists]] \\ ===== Definition of columns ===== ^ name ^ ^ | Code | Name of code | | # of Cores | Number of used CPU cores | | # of Processes | Number of MPI processes | | # of Threads | Number of threads for each process | | $N_{c}$ | Truncation lavel for Chebyshev polynomials | | $l_{max}$ | Truncation lavel for spherical harmonnics | | $(N_{r},N_{\theta},N_{\phi})$ | Number of grids in spherical coordinate | | Elapsed | Elapsed (wall clock time) for one time step | | Nonlinear | Elapsed (wall clock time) for evaluation of nonlinear terms | | Solver | Elapsed (wall clock time) for linear solver (including communications) | | Efficiency | Parallel efficiency | | SUs | Service unit for $10^{4}$ time steps (Core hours) | ===== Single processors Result ===== === Spherical harmonics expansion with Chebyshev expansion === ^ Code ^ $N_{c}$ ^ $l_{max}$ ^ $(N_{r},N_{\theta},N_{\phi})$ ^ Elapsed ^ Nonlinear ^ Solver ^ Communication ^ Note ^ | H2000 | 48 | 47 | ( 72,72,144) | 1.38531 | | | | | | MagIC | 48 | 47 | ( 72,72,144) | 0.35723 | | | | | | GFD-Dennou | 48 | 42 | (48,64,128) | 0.445464 | 0.233546 | 0.183094 | | | | Simitev-Busse-drs | 47 | 42 | (48,64,128) | 0.460322 | 0.378743 | 0.0565741 | 0.005765 | | === Spherical harmonics expansion with FDM === ^ Code ^ $l_{max}$ ^ $(N_{r},N_{\theta},N_{\phi})$ ^ Elapsed ^ Nonlinear ^ Solver ^ Communication ^ Note ^ | Calypso V.1.1 | 47 | (73,72,144) | 1.604797 | 1.56274 | 0.04205857 | 0.508469 | | | Calypso (Latest) | 47 | (73,72,144) | 0.839953 | 0.797941 | 0.0399720 | 0.04934 | | | Parody | 47 | ( 73,72,144) | 0.269091 | 0.257912 | 0.00424973 | | Use 4 MPI threads | | ETH | 47 | (72,72,144) | 1.049 | 0.880913 | 0.168083 | 0.360225 | | | LSD | 48 | ( 73,72,144) | 0.774053 | 0.647997 | 0.647997 | | Nonlinear terms are evaluated twice | | xshells | 47 | ( 73,72,144) | 0.26967 | | | | | === Spherical harmonics expansion with Compact FDM === ^ Code ^ $l_{max}$ ^ $(N_{r},N_{\theta},N_{\phi})$ ^ Elapsed ^ Nonlinear ^ Solver ^ Communication ^ Note ^ | TITECH | 47 | (73,72,144) | 0.678760 | 0.488277 | 0.190479 | 0.029903 | Nonlinear terms are evaluated twice | {{wg:dynamo:Performance_results:comp:comp_single.png?480}}\\ Elapsed (wall clock) time with single core. ===== 256 Processes results ===== === Spherical harmonics expansion with Chebyshev expansion === ^ Code ^ $N_{c}$ ^ $l_{max}$ ^ $(N_{r},N_{\theta},N_{\phi})$ ^ Elapsed ^ Nonlinear ^ Solver ^ Communication ^ Note ^ | UCSC | 255 | 170 | (257,256,512) | 1.33908 | 1.06534 | 0.273739 | 0.080553 | used 257 Cores | | UCSC_latest | 255 | 511 | (257,768,1536) | 8.6780 | | | | used 257 Cores | | MagIC5 | 255 | 170 | (257,256,512) | 0.213723 | 0.120045 | 0.062081 | 0.0240802 | 4 SMP threads | | MagIC5 | 511 | 256 | (513,384,763) | 2.22868 | 0.850126 | 0.442441 | 0.129714 | 8 SMP threads | | GFD-Dennou | 64 | 170 | (64,256,512) | 8.5006 | 3.08209 | 5.17536 | | === Spherical harmonics expansion with FDM === ^ Code ^ $l_{max}$ ^ $(N_{r},N_{\theta},N_{\phi})$ ^ Elapsed ^ Nonlinear ^ Solver ^ Communication ^ Note ^ | Calypso V.1.1 | 255 | (513,384,768) | 6.74711 | 6.69947 | 0.0476359 | 0.72289 | 4 SMP threads | | Calypso (Latest) | 255 | (513,384,768) | 2.30595 | 2.25802 | 0.0409069 | 0.632535 | 4 SMP threads | | Parody | 255 | (512,384,768) | 0.951109 | 0.779863 | 0.122069 | | 8 SMP threads | | ETH | 255 | (512,384,768) | 5.49180 | 5.33942 | 0.152376 | 2.2764 | | | LSD | 256 | (512,384,768) | 4.08026 | 3.85867 | 0.0918966 | | Nonlinear terms are evaluated twice | | xshells | 256 | (512,384,768) | 0.30811 | 0.2502 | 0.02765 | 0.0302 | | === Spherical harmonics expansion with Compact FDM === ^ Code ^ $l_{max}$ ^ $(N_{r},N_{\theta},N_{\phi})$ ^ Elapsed ^ Nonlinear ^ Solver ^ Communication ^ Note ^ | TITECH | 255 | (256,384,768) | 1.86101 | 1.7352 | 0.125808 | 0.475738 | Nonlinear terms are evaluated twice | === FEM === ^ Code ^ $N_{r}$ ^ $N_{sph}$ ^ Elapsed time ^ Solver time ^ Comm. time ^ Note ^ | GeoFEM | 65 | 31106 | 12.0594 | 11.5573 | 1.29915 | 4 SMP threads | {{wg:dynamo:Performance_results:comp:comp_256cores.png?480}}\\ Elapsed (wall clock) time with 256 cores. [[wg:dynamo:Performance_results|Back to performance benchmark lists]] \\