Benchmarks used for my 2015 ALCC proposal

You can create and build a benchmark in one of two ways:

• Manualy, with pattern-matching, genbox, genmap, and makenek
• Automagically, with genrun and makenek

1. Find an existing benchmark that contains a .usr, .rea, .box, and SIZE file
2. Change nelx, nely, nelz in the .box file
3. Change lelg and lp in the SIZE file
4. my_shape in the .usr file to match {nelx, nely, nelz}
5. Change p65 in the .rea file to be some factor of lp
6. Run genbox to update mesh information in the .rea input file
7. Run genmap to create a .map input file
8. Run makenek to build the executable

1. Copy any of the .json files into a new directory
2. Edit procs, shape_mesh, and io_files to get the right scale
3. clone or download maxhutch/nek-tools
4. Run genrun.py -d *.json -u ../iturb_f77.tusr --map --legacy my_benchmark