The OpenCHAMI Developer Summit 2025, hosted at the Texas Advanced Computing Center (TACC), brought together developers, system administrators, and industry partners from across the HPC and AI ecosystem. Over three days (Sept 9–11), the community engaged in strategic discussions, technical deep dives, and hands-on tutorials that advanced OpenCHAMI’s role as the open, community-driven foundation for HPC system management.
Key Themes
- Community & Collaboration – Strong emphasis on open governance, Requests for Discussion (RFDs), and preparing for future hackathons.
- Composable Infrastructure – Presentations on OpenCHAMI’s architecture highlighted sysadmin-first design, fast boot speeds, and deployment at leading HPC centers.
- Scaling & Performance – In-depth discussions focused on scaling to 6,000+ nodes, optimizing boot workflows, and addressing bottlenecks in DHCP and image servers.
- Security & Workflow Improvements – Talks on software supply chain security (Witness by TestifySec) and GitHub workflow automation underscored the community’s focus on reliability and developer productivity.
- Hands-On Tutorials – Participants gained practical experience with OpenCHAMI installation, configuration, image building, and troubleshooting. Tutorials will be repurposed into a YouTube series for broader accessibility.
- Future Planning – Deliverables include visualization clusters by December, new automation for upgrades and artifacts, and deeper community engagement at HPC Asia 2026 and the next Dev Summit (planned with HPSF in Chicago, March 2026).
Help us make future events even better by sharing your feedback with us.
We’re pleased to share a recap of the proceedings, along with recordings of many of the talks.
📌 Day 1 Highlights – Foundations & Vision
- Welcome & Community Building – Alex Lovell-Troy emphasized collaboration, tutorials, hackathons, and governance discussions.
- KeyNote – System Management at TACC – Dan Stanzione discussed managing national-scale systems, automation evolution, and upcoming Grace Hopper/VERA nodes. 🎥 Recording.
- Dell’s Open-Source HPC Strategy – Onur Celebioglu highlighted Dell’s adoption of open standards, Omnia, and contributions to OpenCHAMI. 🎥 Recording.
- OpenCHAMI Composable Infrastructure – Alex Lovell-Troy presented sysadmin-first design, boot speed gains, and multi-site deployments.
- Site & Tool Updates – Progress from NERSC, CSCS, Bristol, and TACC; demo of Virtual Test Development System (vTDS).
- Key Technical Topics – Cloud-init security, secrets management, DHCP scaling, and logging standardization.
📌 Day 2 Highlights – Tutorials & Secure Workflows
- Why OpenCHAMI Matters – Adam Hough on the need for a community-driven booting standard; HPE and Broadwing confirmed support. 🎥 Recording.
- Hands-On Tutorials (Devon Bautista) – Covered installation, architecture, boot configuration, layered image builds, and ARM considerations.
- Supply Chain Security – Cole Kennedy (TestifySec) demoed Witness for verifying builds and SBOMs. 🎥 Recording
- GitHub Workflow Optimization – Taylor (Broadwing) presented automation strategies for CI/CD and repo organization. 🎥 Recording
📌 Day 3 Highlights – Scaling & Hackathon Prep
- Systems Architecture & Kernel Boot Demo – Nick showcased OpenCHAMI’s boot process, lifecycle management, and scaling to 6,000+ nodes. 🎥 Recording
- Hackathon Planning – Teams formed around Image Builder improvements, GitHub Actions, and CI/CD testing.
- Technical Challenges – Discussions on scaling DHCP/image servers, reducing boot times, and artifact deployment automation.
- Community Updates – Architectural decisions, GitHub team changes, dev build improvements, and planning for future summits.
- Looking Ahead – SC25 BoF hosted by HPE and Dell, Visualization clusters by December, tutorials to YouTube, participation at HPC Asia 2026, and next Dev Summit in Chicago (March 2026).
🙌 Thank You
We are grateful to all speakers, organizers, and participants for making this summit a success. Your continued contributions are what make OpenCHAMI thrive.
Stay tuned for updates on hackathon outcomes, upcoming tutorials, and opportunities to get involved in RFDs and community discussions.
✅ Takeaway: The Summit established OpenCHAMI as the collaborative backbone for modern HPC and AI infrastructure management, combining technical innovation with a vibrant and engaged community.
See you on Slack and at the next summit!
— The OpenCHAMI Team