Held virtually September 29 - 30, 2023, the NERSC User Group (NUG) Annual Meeting is an opportunity for the entire NERSC community to gather to share information and research achievements.
This event is organized by NERSC Staff and the NUG Executive Committee.
The Call for Presentations is now open!
Please view the topics and submission guidelines listed below before preparing your submission.
The deadline to submit a presentation abstract for consideration is:
5pm PDT on Monday, August 24th, 2026
This year's NERSC User Group (NUG) Annual Forum will include two forms of short talks from NERSC users, in order to illustrate the work done on Perlmutter.
Presentation options include:
12 minute (10 for presenting + 2 for questions) presentation to be presented live during the event
5 minute pre-recorded presentations to be featured on the meeting website
Early career users, such as undergraduate and graduate students, as well as postdocs and new PIs, are particularly encouraged to submit talks.
Each submission must include which major topic area it falls into (numbered items), as well as the sub-topic within the major topic (bulleted items). You may also provide key words specific to your work.
1. Heterogeneous and Parallel HPC
- Performance modeling and optimization
- Application development
- Containerization
- Resilience and fault tolerance at scale
2. Applied AI and ML
- AI-accelerated code development and/or maintenance
- Agentic AI usage for experiment or workflows
- Autonomous experimentation
3. HPC Workflows
- Cross-facility workflows
- Application workflows
- CI/CD at NERSC
- API-based NERSC usage
- Software sustainability and packaging
4. Data Management, Storage, and I/O
- Parallel I/O and storage system optimization
- Archival storage and data lifecycle management
- Large-scale data transfer and networking
- Visualization and scientific data analysis
- Reproducibility, provenance, and FAIR data
5. HPC User Experience
- New user experience and onboarding
- Meta-analysis of NERSC usage
- HPC education and training
- Broadening participation and workforce development
6. Performance Optimization and Scaling for New HPC Architectures
- AI-acceleration
- Energy-efficient computing
7. Edge Computing and Facility Integration
- Data streaming to HPC
- Edge, realtime, and/or low-latency computing
- Experimental and observational facility integration
8. Security and Trusted Computing
- Secure computing for sensitive/controlled-access data (e.g., genomic, health data)
- Identity, access, and authentication workflows
9. Quantum Computing
- Algorithm development
- Hybrid quantum-classical workflows
To ensure a strong, well-rounded set of talks, the following criteria will be used to evaluate proposals:
Use of NERSC resources. How clearly does the work highlight a novel and impactful use of NERSC resources?
Relevance to the broader user community. Does the talk offer lessons, techniques, or takeaways other NERSC users could actually apply, rather than being of interest only to a narrow subfield?
Collaboration between institutions. Does the talk highlight team science and collaboration between research institutions?
Clarity and quality of the abstract. Is the proposed talk well-scoped and clearly aligned with the chosen topic area? This is important so attendees can easily find and engage with presentations that fall within their area of interest.
Relevance to the HPC community. Is this recent or ongoing work? Is the work novel and has not been presented in the past?
Accessibility to a general HPC audience. Even if the talk is technical, can it be understood by NERSC users outside that specific domain?