SingularityCE

What SingularityCE is

SingularityCE is a free, open-source container tool from Sylabs, Inc. It packages software into a single file designed for portable, reproducible use. Created for complex applications on high-performance computing (HPC) clusters, it is also used in academia and industry. A container built on a laptop can run on HPC clusters, university or company clusters, servers, cloud environments, or workstations. SingularityCE uses the single-file Singularity Image Format and supports compatibility with Docker and OCI container formats.

Users can build containers from definition files, use remote builds, or build as a non-root user with fakeroot. The product page states built-in support for GPUs and high-speed interconnects; its guide also covers GPU access and containerized MPI applications, along with OCI registries and Sylabs Container Services Library usage. Containers can be signed with PGP keys for verification and encrypted with LUKS2 so they can run without first being decrypted to disk. By default, a user inside a container keeps the same identity as outside and cannot gain additional privileges on the host. SingularityCE runs on Linux and is self-hosted. Sylabs says engineering and engineering support will cease on October 25, 2026; it expects version 4.6 to be its final release managed by the company, after which it will not provide security, bug-fix, or feature development.

Who it suits

SingularityCE suits researchers, HPC users, and teams in academia or industry that need to package software for portable use across clusters, servers, cloud environments, and workstations. Its Linux, self-hosted model is aimed at users who can manage their own container environment.

Limits

The free plan is open source and BSD licensed. Sylabs states that engineering and engineering support will cease on October 25, 2026, and that it expects version 4.6 to be its final release managed by the company; after that, it will not provide security, bug-fix, or feature development.

Getting started

1. Visit the SingularityCE website and its guide.
2. Prepare a Linux host for the self-hosted deployment.
3. Build an image from a definition file, use a remote build, or build with fakeroot.
4. Run the resulting Singularity Image Format container on a supported environment.
5. Use the guide for GPU access, MPI applications, OCI registries, or Sylabs Container Services Library workflows as needed.

Verdict

Choose SingularityCE for free, Linux-based container packaging across HPC and other computing environments, especially when GPU support, MPI workloads, or rootless building matter. Its portable image format and Docker and OCI compatibility support varied workflows. Look elsewhere if you need Windows containers or ongoing engineering and engineering support from Sylabs after October 25, 2026.