Mixed-precision Nekbone

Nekbone solves a standard Poisson equation using a conjugate gradient iteration with a simple or spectral element multigrid preconditioner on a block or linear geometry. It exposes the principal computational kernel to reveal the essential elements of the algorithmic-architectural coupling that is pertinent to Nek5000. Original repository: https://github.com/Nek5000/Nekbone

It's an implementation of mixed-precision Nekbone.

Usage

Supported compilers

To compile mixed-precision Nekbone, both Fortran compiler and C compiler is required. Tested compilers include:

• Classic Flang: flang & clang (https://github.com/flang-compiler/flang)
• Intel compiler: ifort & icc
• Intel oneAPI compiler: ifx & icx
• Cray compiler: ftn & cc
• GNU compiler: gfortran & gcc (old version like gfortran-5)

Run test cases

• Test case w/o preconditioner: test/nek
• Test case w preconditioner: test/mgrid

Steps:

1. Specify compilers in makenek file: F77 and CC, flang/clang is default.
2. (Optional) For parallel building: comment IFMPI="false" and specify MPI compiler.
3. (Optional) For changing optimization flags: modify OPT_FLAGS_STD and OPT_FLAGS_MAG.
4. Run the script: ./makenek
5. Define the parameters for simulation in data.rea.
6. Run the program ./nekbone.