1. Code Name: XGC1
2. Code Category: Kinetic edge Turbulence
3. Primary Developer: S. Ku and C.S. Chang
4. Other Developers and Users: S. Klasky, P. Worley, E. D'Azevedo (ORNL), M. Adams (Columbia U.), J. Cummings (Caltech), G. Diff (UCSD), S. Parker (U. Colorado), Z. Lin (UCI)
5. Short description (one line if possible): Gyrokinetic particle-in-cell code in diverted magnetic geometry with wall boundary condition.
6. Computer Language (Fortran77, Fortran90, C, C++, etc) and approx # of lines: Fortran90, approximately 30,000 lines:
7. Type of input required (including files and/or output from other codes). Is there any special input preparation system (eg, GUI): EFIT geqdsk file or equivalent, grid data file from a small code included in XGC-1 package and the TRIANGLE code.
8. Type of output produced (including files that are read by other codes and sizes of large files and synthetic diagnostics): bp files (ADIOS output, easily converted to HDF5 or NETCDF). Plasma profile data (~ 100 MB), turbulence data (~10 GB), restart file (~1 TB)
9. Describe any postprocessors which read the output files: ADIOS utilities for converting to hdf5. Matlab.
10. Status and location of code input/output documentation: Basic documentation exists at http://w3.physics.lehigh.edu/~xgc/ 11. Code web site? http://w3.physics.lehigh.edu/~xgc/
12. Is code under version control? What system? Is automated regression testing performed? Code is in SVN. Automated regression testing not performed.
13. One to two paragraph description of equations solved and functionality including what discretizations are used in space and time: XGC-1 is Particle-in-cell code based upon LittleJohn-Boozer Lagrangian Equation of Motion. R-K or predictor-corrector is used for time advance. Field data is discretized with unstructured triangular FEM grid.
14. What modes of operation of code are there (eg: linear, nonlinear, reduced models, etc ): Full-f or delta-f. Linear Monte-Carlo or nonlinear PDE collision operations.
15. Journal references describing code: S. Ku, C.S. Chang, and P.H. Diamond, Nuclear Fusion 49, 115021 (2009); C.S. Chang,, S. Ku, P. H. Diamond, Z. Lin, S. Parker, T. S. Hahm and N. Samatova, "Compressed ion temperature gradient turbulence in diverted tokamak edge," Phys. Plasmas 16, 056108 (2009)
16. Codes it is similar to and differences (public version): Turbulence version of XGC0
17. Results of code verification and convergence studies (with references): S. Ku, C.S. Chang, and P.H. Diamond, Nuclear Fusion 49, 115021 (2009)
18. Present and recent applications and validation exercises (with references as available): Edge ITG turbulence study [C.S. Chang,, S. Ku, et al, Phys. Plasmas 16, 056108 (2009)]. Core-edge nonlocal ITG interaction and global self-organization [S. Ku, C.S. Chang, and P.H. Diamond, Nuclear Fusion 49, 115021 (2009)], Full-f momentum transport by ITG [S. Ku, et al, IAEA2010]
19. Limitations of code parameter regime (dimensionless parameters accessible) : Wave number k rho_i should be 1 or less , perturbation level delta_n / n > 1%
20. What third party software is used? (eg. Meshing software, PETSc, ...): Commercial software is excluded for portability. PSPLINE, ADIOS, PETSc, TRIANGLE
21. Description of scalability: Code is proven to scale efficiently to the maximal available Jaguarpf (to 223,488 cores).
22. Major serial and parallel bottlenecks: serial - particle motions within a core, parallel - field data communication and field solver.
23. Are there smaller codes contained in the larger code? Grid generation code is included.
24. Supported platforms and portability: Jaguar and Franklin (Cray XT-5). Can be portable to other platforms
25. Illustrations of time-to-solution on different platforms and for different complexity of physics, if applicable: One day run is aimed for most of the realistic geometry simulations.