We’re excited to welcome Ramble to the High Performance Software Foundation!
Ramble is an open source experimentation framework written in Python, focused on capturing provenance information and improving the reproducibility of computational experiments. It allows users to rapidly create, execute, analyze, and manage large sets of experiments across HPC, AI, and ML environments.
The framework supports a wide range of use cases, including test suites, parameter studies, performance tuning explorations, and acceptance testing. Ramble can also automate several stages of an experiment’s lifecycle, including downloading input files, creating software stacks through package managers such as Spack, executing experiments, extracting figures of merit, and archiving results.
Reproducibility has become an increasingly important topic across the high performance computing community, including at major industry events such as SC Conference. Ramble’s focus on reproducible experimentation helps researchers and developers better validate, share, and scale their work across systems and teams.
“Ramble was created to make computational experiments easier to reproduce, manage, and scale across a variety of HPC and AI environments,” said the creator of Ramble, Doug Jacobsen, Software Engineer, Google LLC. “By joining the High Performance Software Foundation, we’re excited to collaborate more closely with the broader open source community and continue improving accessibility for users working across HPC, AI, and ML workloads.”
What Makes Ramble Different
Ramble focuses on making experimentation workflows easier to manage from end to end. Instead of treating experiments as one off executions, Ramble helps users standardize and automate the entire process, from environment setup through analysis and archival.
The project supports integrations with multiple software environment and package management systems, including Slurm, Spack, environment modules, EESSI, and pip. This flexibility allows researchers and developers to adapt workflows across different infrastructures while maintaining reproducibility and consistency. The abstraction layers in Ramble allow users to easily compose portions of their experiments together, which can make it easier to adapt to new systems and workflows or to explore how changing out a specific layer impacts the overall experiment.
Ramble is currently operated as a community driven project on GitHub, with contributors spanning multiple institutions and technical communities. The project is committed to supporting users working across traditional HPC workloads as well as emerging AI and ML applications.
What’s Next
As part of the High Performance Software Foundation, Ramble plans to continue growing its contributor community, improving usability, and expanding support for reproducible experimentation workflows across HPC, AI, and ML ecosystems.
Future efforts will focus on streamlining experiment management, improving integrations with surrounding infrastructure tooling, and making it easier for users to adopt reproducible workflows at scale.
“Open source software is critical for reproducible science,” said Todd Gamblin, HPSF Governing Board Chair. “Projects like Ramble help researchers and developers build reliable, reproducible experiments and benchmarks. We’re happy to have Ramble in HPSF, where we can help to lower barriers to adoption in HPC and AI and ultimately grow the Ramble community.”
Learn More
To learn more about Ramble, explore the project’s documentation and community resources to see how it supports reproducible computational experimentation across HPC, AI, and ML workflows.
Get Involved
The Ramble community welcomes contributors, users, and collaborators from across the ecosystem. Whether you’re building HPC applications, managing AI experiments, or improving reproducibility practices in research computing, there’s a place to contribute and help shape the future of open source experimentation workflows.