| 710 | | |
| 711 | | |
| 712 | | ---- |
| 713 | | |
| 714 | | |
| 715 | | = Nbody Benchmark (MPI+!OmpSs-2+TAMPI) = |
| 716 | | |
| 717 | | Users must clone/download this example's repository from [https://pm.bsc.es/gitlab/ompss-2/examples/nbody] and transfer it to a DEEP working directory. |
| 718 | | |
| 719 | | == Description == |
| 720 | | |
| 721 | | This benchmark represents an N-body simulation to numerically approximate the evolution of a system of bodies in which each body continuously interacts with every other body. A familiar example is an astrophysical simulation in which each body represents a galaxy or an individual star, and the bodies attract each other through the gravitational force. |
| 722 | | |
| 723 | | There are **7 implementations** of this benchmark which are compiled in different |
| 724 | | binaries by executing the command `make`. These versions can be blocking, |
| 725 | | when the particle space is divided into smaller blocks, or non-blocking, when |
| 726 | | it is not. |
| 727 | | |
| 728 | | The interoperability versions (MPI+!OmpSs-2+TAMPI) are compiled only if the environment variable `TAMPI_HOME` is set to the Task-Aware MPI (TAMPI) library's installation directory. |
| 729 | | |
| 730 | | == Execution Instructions == |
| 731 | | |
| 732 | | The binaries accept several options. The most relevant options are the number |
| 733 | | of total particles (`-p`) and the number of timesteps (`-t`). More options |
| 734 | | can be seen with the `-h` option. An example of execution could be: |
| 735 | | |
| 736 | | `mpiexec -n 4 -bind-to hwthread:16 ./nbody -t 100 -p 8192` |
| 737 | | |
| 738 | | in which the application will perform 100 timesteps in 4 MPI processes with 16 hardware threads in each process (used by the !OmpSs-2 runtime). The total number of particles will be 8192 so that each process will have 2048 particles (2 blocks per process). |
| 739 | | |
| 740 | | == References == |
| 741 | | |
| 742 | | * [https://pm.bsc.es/gitlab/ompss-2/examples/nbody] |
| 743 | | * [https://en.wikipedia.org/wiki/N-body_simulation] |
| 744 | | |
| 745 | | |
| 746 | | ---- |
| 747 | | |
| 748 | | |
| 749 | | = Heat Benchmark (MPI+!OmpSs-2+TAMPI) = |
| 750 | | |
| 751 | | Users must clone/download this example's repository from [https://pm.bsc.es/gitlab/ompss-2/examples/heat] and transfer it to a DEEP working directory. |
| 752 | | |
| 753 | | == Description == |
| 754 | | |
| 755 | | This benchmark uses an iterative Gauss-Seidel method to solve the heat equation, |
| 756 | | which is a parabolic partial differential equation that describes the distribution of heat (or variation in temperature) in a given region over time. The heat equation is of fundamental importance in a wide range of science fields. In |
| 757 | | mathematics, it is the parabolic partial differential equation par excellence. In statistics, it is related to the study of the Brownian motion. Also, the diffusion equation is a generic version of the heat equation, and it is related to the study of chemical diffusion processes. |
| 758 | | |
| 759 | | There are **9 implementations** of this benchmark which are compiled in different |
| 760 | | binaries by executing the command `make`. |
| 761 | | |
| 762 | | The interoperability versions (MPI+!OmpSs-2+TAMPI) are compiled only if the environment variable `TAMPI_HOME` is set to the Task-Aware MPI (TAMPI) library's installation directory. |
| 763 | | |
| 764 | | == Execution Instructions == |
| 765 | | |
| 766 | | The binaries accept several options. The most relevant options are the size |
| 767 | | of the matrix in each dimension (`-s`) and the number of timesteps (`-t`). More options can be seen with the `-h` option. An example of execution |
| 768 | | could be: |
| 769 | | |
| 770 | | `mpiexec -n 4 -bind-to hwthread:16 ./heat -t 150 -s 8192` |
| 771 | | |
| 772 | | in which the application will perform 150 timesteps in 4 MPI processes with 16 |
| 773 | | hardware threads in each process (used by the !OmpSs-2 runtime). The size of the |
| 774 | | matrix in each dimension will be 8192 (8192^2^ elements in total), this means |
| 775 | | that each process will have 2048x8192 elements (16 blocks per process). |
| 776 | | |
| 777 | | == References == |
| 778 | | |
| 779 | | * [https://pm.bsc.es/gitlab/ompss-2/examples/heat] |
| 780 | | * [https://pm.bsc.es/ftp/ompss-2/doc/examples/local/sphinx/04-mpi+ompss-2.html] |
| 781 | | * [https://en.wikipedia.org/wiki/Heat_equation] |
| 782 | | |
| 783 | | ---- |
| 784 | | |
| 785 | | = Krist Benchmark (!OmpSs-2+CUDA) = |
| 786 | | |
| 787 | | Users must clone/download this example's repository from [https://pm.bsc.es/gitlab/ompss-2/examples/krist] and transfer it to a DEEP working directory. |
| 788 | | |
| 789 | | == Description == |
| 790 | | |
| 791 | | This benchmark represents the krist kernel, which is used in crystallography to find the exact shape of a molecule using Rntgen diffraction on single crystals or powders. |
| 792 | | |
| 793 | | There are **2 implementations** of this benchmark, ''krist'' and ''krist-unified'' using regular and unified CUDA memory, repectively. |
| 794 | | |
| 795 | | == Execution Instructions == |
| 796 | | |
| 797 | | `./krist N_A N_R` |
| 798 | | |
| 799 | | where: |
| 800 | | * `N_A` is the number of atoms (1000 by default). |
| 801 | | * `N_R` is the umber of reflections (10000 by default). |
| 802 | | |
| 803 | | == References == |
| 804 | | |
| 805 | | * [https://pm.bsc.es/gitlab/ompss-2/examples/krist] |
