This BoF brings together a panel of international HPC software development experts to discuss tools and techniques for enhancing HPC software developer productivity. Key themes include DevOps, programming models (OpenMP, RAJA, Kokkos, etc.), and development tools. Additionally, the integration of AI across these themes will be explored (e.g., is GitHub Copilot useful in HPC software development?). Attendees will gain insights into overcoming productivity challenges unique to HPC. This session is ideal for HPC software developers seeking to improve their productivity through innovative tools and shared expert experiences. LLNL-ABS-866156.
Our HPC developer productivity survey results here!
Logistics
Our BoF will take place from 5:15pm to 6:45pm EST on Tuesday, 11/19/2024 at SC24. For up-to-date information regarding the room that the BoF will take place in, see the SC24 schedule.
The BoF will consist of a panel and a signficant amount of time dedicated to audience Q&A.
Slides and survey results from this BoF will be posted on this website after the BoF is held.
Panelists
Ivy Peng
Ivy Peng is an Associate Professor of Computer Science at KTH Royal Institute of Technology in Sweden. She leads the Scalable Computing Laboratory (ScaLab) at EECS school. Her research interests include heterogeneous memory systems, near-memory computing, persistent memory, and disaggregated memory. Ivy Peng has authored over 60 peer-reviewed publications at various conferences and journals. Ivy Peng is the Coordinator of EU Horizon Project OpenCUBE on Open-Source Cloud-Based Services on EPI Systems and the PI of several Swedish national research projects. Before her tenure at KTH, she was a Computer Scientist at Lawrence Livermore National Laboratory in USA.
Chris White
Chris White is Lawrence Livermore National Lab’s WSC DevOps Coordinator, focusing on program-wide DevOps efforts, such as unifying and improving build and automation processes and common tools, such as BLT, a streamlined CMake-based foundation for Building, Linking, and Testing large-scale high-performance computing (HPC) applications. He also leads the CS efforts on Serac, a high-order nonlinear thermomechanical simulation code, advising a multi-disciplinary team on software best practices. In addition, Chris is a Group Lead, supervising, hiring, and mentoring a team of computer scientists. He graduated from UC Santa Cruz with a BS in Computer Science and has over 17 years of experience, including working at Borland Software, where he focused on streamlining and hardening DevOps for their IDE Tools division.
Cameron Rutherford
Cameron Rutherford is a Research Software Engineer at Pacific Northwest National Laboratory (PNNL) in the HPC/ML/Q group, and Adjunct Professor at Fordham University. He led the ExaSGD project as Software Lead from 2021 to 2023, deploying GitHub and GitLab pipelines and contributing to the largest grid simulation to date. His work at PNNL also involves internal DevOps and MLOps initiatives, including containerization, CI/CD, privacy-preserving ML, and LLM benchmarking. As an experienced HPC developer, he has ported Fortran code to C++ for E3SM and Eagles, implemented CI/CD for machine learning projects, and provided “Help Desk” support across PNNL. https://fordh.am/1p3.
Adrian Jackson
Adrian Jackson is a researcher into high performance computing at EPCC, The University of Edinburgh. He has spent 2+ decades working with computational scientists on codes that have often resembled the worst nightmare of a software engineer, but that produce unrivalled science. Fortran has been a constant companion, and CMake a frequent nemesis, but tools for understanding both code functionality and code performance have definitely been what has made life bearable during these endeavors. He has worked across domains such as Plasma Physics and Fusion, to Computational Fluid Dynamics, to Radio Astronomy and Evolutionary Biology. When not wrangling codes he also researchers novel HPC hardware, with a particular focus on memory, and does a lot with filesystems and other storage approaches.