Section author: Axel Huebl

Installing PIConGPU means installing C++ libraries that PIConGPU depends on and setting environment variables to find those dependencies. The first part is usually the job of a system administrator while the second part needs to be configured on the user side.

Depending on your experience, role, computing environment and expectations for optimal hardware utilization, you have several ways to install and select PIConGPU’s dependencies. Choose your favorite install and environment management method below, young padawan, and follow the corresponding sections of the next chapters.

Ways to Install

Choose one of the installation methods below to get started.

Load Modules

On HPC systems and clusters, software is usually provided by system administrators via a module system (e.g. [modules], [Lmod]). In case our software dependencies are available, we usually create a file in our $HOME named <queueName>_picongpu.profile. It loads according modules and sets helper environment variables.


For many HPC systems we have already prepared and maintain an environment which will run out of the box. See if your system is in the list so you can skip the installation completely!


[Spack] is a flexible package manager that can build and organize software dependencies. It can be configured once for your hardware architecture to create optimally tuned binaries and provides modulefile support (e.g. [modules], [Lmod]). Those auto-build modules manage your environment variables and allow easy switching between versions, configurations and compilers.

Build from Source

You choose a supported C++ compiler and configure, compile and install all missing dependencies from source. You are responsible to manage the right versions and configurations. Performance will be ideal if architecture is chosen correctly (and/or if built directly on your hardware). You then set environment variables to find those installs.


We currently do not have an official conda install (yet). Due to pre-build binaries, performance could be not ideal and HPC cluster support (e.g. MPI) might be very limited. Useful for small desktop or single-node runs.


Not yet officially supported [nvidia-docker], but we already provide a Dockerfile to get started. Performance might be not ideal if the image is not built for the specific local hardware again. Useful for small desktop or single-node runs. We are also working on Singularity images.



T. Gamblin and contributors. A flexible package manager that supports multiple versions, configurations, platforms, and compilers, SC ‘15 Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (2015), DOI:10.1145/2807591.2807623,


J.L. Furlani, P.W. Osel. Abstract Yourself With Modules, Proceedings of the 10th USENIX conference on System administration (1996),


R. McLay and contributors. Lmod: An Environment Module System based on Lua, Reads TCL Modules, Supports a Software Hierarchy,


Nvidia Corporation and contributors. Build and run Docker containers leveraging NVIDIA GPUs,