Getting Started

Requirements

• NVIDIA graphics card
• C/C++ compiler, ex. g++
• NVIDIA cuda compiler (the CUDA toolkit is available here)
• hdf5 library (recommended)

Downloading the code

The public version of the code can be found at https://github.com/cholla-hydro/cholla. To download it, you can either clone the main repository directly

git clone https://github.com/cholla-hydro/cholla

or create your own fork on github and clone that (recommended if you plan to contribute).

Compiling the code

The main repository contains two makefiles, one for linux and one for macs. Once you have downloaded the required compilers and libraries, you should be able to compile the code by selecting the relevant makefile

make -f makefile.linux

If successful, this should create an executable called "cholla" in the main directory. You may have to edit the makefile to tell it where the libraries are. If you are running cholla on a cluster, you may also need to load the relevant modules prior to compiling the code.

Running the code (serial mode)

To run cholla on a single GPU, you must execute the binary and provide it with an input parameter file. For example, to run a 1D Sod Shock tube test, you would run:

./cholla tests/1D/Sod.txt

The code will write some information about the input parameters to the terminal:

Parameter values:  nx = 256, ny = 1, nz = 1, tout = 0.200000, init = Riemann, boundaries = 3 3 0 0 0 0
Output directory:  ./
Local number of grid cells: 256 1 1 262

followed by some text indicating the code is initializing:

Setting initial conditions...
Setting boundary conditions...
Boundary conditions set.
Dimensions of each cell: dx = 0.003906 dy = 0.003906 dz = 0.003906
Ratio of specific heats gamma = 1.400000
Nstep = 0  Timestep = 0.000000  Simulation time = 0.000000
Writing initial conditions to file...
Starting calculations.

After this, the code will print out a line for every time step it takes, indicating the step it is on, the total time elapsed in the simulation, the size of the timestep taken, the wall-time elapsed during the timestep, and the total wall-time of the simulation:

n_step: 1   sim time:  0.0009904   sim timestep: 9.9042e-04  timestep time =   301.153 ms   total time =    0.3060 s

The code will stop running when it reaches the final time specified in the input file.