Julia¶
Julia is a high-level, high-performance dynamic programming language for technical computing. It has syntax that is familiar to users of many other technical computing environments. Designed at MIT to tackle large-scale partial-differential equation simulation and distributed linear algebra, Julia features a robust ecosystem of tools for optimization, statistics, parallel programming, and data visualization. Julia is actively developed by teams at MIT and in industry, along with hundreds of domain-expert scientists and programmers from around the world.
Using Julia at NERSC¶
Julia 1.4.2 and 1.6.0 are available at NERSC by loading the julia module:
module load julia/1.6.0
This also configures the JULIA_DEPOT_PATH and JULIA_LOAD_PATH to give precedence to packages installed in the global environment.
We provide the following packages:
which which are configured specifically for NERSC's systems. They have been installed in the global environment globalenv.
Any packages you install using Pkg.add will be installed to the user's local julia package directory ~/.julia.
Running Pkg.add Will Overwrite Pre-Configured NERSC Packages
If you run Pkg.add for any packages that are already installed in globalenv, then Pkg will create a local install which clobbers the\ version in globalenv. For convenience packages like BenchmarkTools this is not a problem. But packages that are pre-configured for NERSC's systems (such as MPI.jl), the local copy will not perform well -- if it runs at all.
To request new packages to be added to globalenv please open a support ticket. Packages used by many users, especially those that require specialized configuration to run on NERSC's systems, will be considered to be included in the list above.
Using Julia with MPI¶
The MPI.jl package has been configured to work with srun. For convenience, we also include the mpiexecjl wrapper. Therefore an MPI-enabled Julia program prog.jl can be executed on N ranks either using:
srun -n N julia prog.jl
or
mpiexecjl -n N julia prog.jl
Developing Julia Code for GPUs¶
Please take a look at our Perlmutter transition guide's section on Julia.
