Getting started

The GRChombo code can in principle be run in serial on a desktop machine. However, for most problems of interest one will need to run the code in parallel. Therefore the instructions here describe how to compile the code in parallel, and assume that you are running it on a remote cluster.

Whilst there is no "typical" problem size, as an order of magnitude estimate a run might use 96-128 cores with 4-6GB of memory each (for smaller cores one can use more cores and share the memory) and take a day to run.

The keys steps, which are detailed in this section, are as follows:

1. Check that the prerequisites are available on your machine or cluster
   • Prerequisites
2. Get and compile Chombo
   • Compiling Chombo
3. Get and compile GRChombo
   • Compiling GRChombo
4. Run one or all of the examples
   • Running examples (probably the best one is the BinaryBH example)
5. Visualise the hdf5 output
   • Visualising outputs

Alternative setup using docker : Getting started fast with docker

It is not unusual to have problems with steps 1 to 3, because the compilers can be fussy and vary from machine to machine. Some common issues are addressed in Compilation FAQ.