Welcome to the Julia for High-performance Computing Tutorial

Goal

The goal of the tutorial is to introduce an audience familiar with high-performance computing (HPC) concepts to the Julia programming language using a hands-on approach.

It’s a good time to get the latest Julia installed from their website

Covered:

  • Workflow aspects: simulation CPU/CPU, parallel I/O, data analysis (e.g. Notebooks)
  • Multiple dispatch for array back ends (CPU, GPU: CUDA, AMDGPU)
  • Multithreaded CPU using Julia @threads
  • GPU programming using vendor (CUDA.jl, AMDGPU.jl) and performance portable layer (JACC.jl, KernelAbstractions.jl)
  • Parallel I/O with ADIOS2.jl (outputs can be visualized in ParaView locally)
  • Julia on OLCF’s JupyterHub and Pluto.jl Notebooks (read and plot using Makie.jl)

Not Covered (future topics?):

  • More general aspects of the language (software engineering, AI)
  • AI infrastructure: Flux.jl, SciML
  • Ahead-of-time compilation: PackageCompiler.jl
  • Tools ecosystem for profiling, debugging (well some vendor supported)

Why Julia?

“Can a machine translate a sufficiently rich mathematical language into a suffienciently economical program at a sufficiently low cost to make the whole affair feasible?” – John Backus on Fortran (1980)

  1. Compiled via LLVM for performance. Julia is Compiled
  2. Provides a unified tested ecosystem (packaging, testing, tools, interactivity, etc.). See JuliaCon presentation on GeneralRegistry.
  3. Community of HPC people from day-1 (the community is you, low entry barrier)
  4. Sustained by JuliaHub, MIT’s JuliaLab, NumFocus and several open-source contributors
  5. Great math (Fortran) and data science (Python) simple syntax
  6. Rapid prototyping: stay in one language+ecosystem for your custom kernels + data analysis
  7. Interoperate with Fortran, C, C++ (via C), Python, R code in a lightweight manner
  8. Renewed interest for high-performance + high-productivity languages
  9. Great first language to expose newcomers to both HPC + math concepts
  10. Contribute to HPC stack: MPI.jl, CUDA.jl, AMDGPU.jl, KernelAbstractions.jl, JACC.jl, ADIOS2.jl, HDF5.jl.
  11. Julia is AI agents friendly due to its simplified syntax

Julia’s value proposition

Why not Julia?

  1. AI has made language choice less relevant
  2. Already existing infrastructure/expertise (Fortran is perfect!)
  3. Too much of an investment to learn a new language
  4. Performance/productivity solutions already in place
  5. Having N-language/ecosystems is not a problem
  6. Python + X is well supported for your needs
  7. Garbage collection, Just-in-time (JIT) compilation are issues
  8. Developing low-level system libraries
  9. Domain outside “technical computing”
  10. Overall, you do not see any value added

Resources

  1. Julia’s value proposition for Better Scientific Software. High-level intro and many resources for Scientific Software.

  2. Bridging HPC Communities through the Julia Programming Language

  3. JuliaParallel.org. Checkout the monthly HPC call and previous events: SC24 BoF.

  4. JuliaCon

  5. OLCF September User’s Call presentations

  6. Julia for ORNL Science Workshop

  7. Julia Basics Notebooks. Run on myBinder or locally.

  8. From zero to Julia!. Series of introductory lessons.

  9. Julia for Sustainable HPC Software by Valentin Churavy.

  10. William F. Godoy, Pedro Valero-Lara, T. Elise Dettling, Christian Trefftz, Ian Jorquera, Thomas Sheehy, Ross G. Miller, Marc Gonzalez-Tallada, Jeffrey S. Vetter, and Valentin Churavy. “Evaluating performance and portability of high-level programming models: Julia, Python/Numba, and Kokkos on exascale nodes.” In 2023 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), pp. 373-382. IEEE, 2023. doi:10.1109/IPDPSW59300.2023.00068

  11. William F. Godoy, Pedro Valero-Lara, Caira Anderson, Katrina W. Lee, Ana Gainaru, Rafael Ferreira Da Silva, and Jeffrey S. Vetter. 2023. Julia as a unifying end-to-end workflow language on the Frontier exascale system. In Proceedings of the SC ‘23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis (SC-W ‘23). Association for Computing Machinery, New York, NY, USA, 1989–1999. doi:10.1145/3624062.3624278

  12. Pedro Valero-Lara, William F. Godoy, Het Mankad, Keita Teranishi, Jeffrey S. Vetter, Johannes Blaschke, and Michel Schanen. “JACC: leveraging HPC meta-programming and performance portability with the just-in-time and llvm-based julia language.” In SC24-W: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis, pp. 1955-1966. IEEE, 2024. doi:10.1109/SCW63240.2024.00245

  13. Pedro Valero-Lara, William F. Godoy, Keita Teranishi and Jeffrey S. Vetter, “JACC.shared: Leveraging HPC Metaprogramming and Performance Portability for Computations That Use Shared Memory GPUs,” 2024 IEEE High Performance Extreme Computing Conference (HPEC), Wakefield, MA, USA, 2024, pp. 1-7, doi:10.1109/HPEC62836.2024.10938453.

  14. Pedro Valero-Lara, William F. Godoy, Philip W. Fackler, Keita Teranishi and Jeffrey S. Vetter, “Enabling Scientific Applications with Performance-Portability and High-Productivity for Multi-GPU Programming with JACC.Multi,” 2025 IEEE International Conference on eScience (eScience), Chicago, IL, USA, 2025, pp. 240-249, doi:10.1109/eScience65000.2025.00036

Acknowledgements

This tutorial is supported by the following ASCR projects: S4PST of the Next Generation of Scientific Software Technologies, and Fairbanks/MAGMA part of competitive portfolios. The many people in the Julia community that made this possible. Thanks to Suzanne Parete-Koon and MacKenzie M. Boyd from the Oak Ridge Leadership Computing Facility (OLCF).


This site uses Just the Docs, a documentation theme for Jekyll.