Nuclear and elementary particle physics increasingly rely on high-performance, composable software stacks to model complex systems, analyze large datasets, and prototype experimental workflows. Julia enables a unique convergence of scientific modeling, performance portability, and long-term maintainability that is difficult to reproduce in Python/C++ ecosystems.
JuliaHEP Mini 2026 presents concrete scientific case studies in nuclear physics and particle/astroparticle physics that were developed end-to-end in Julia. These range from experimental simulation to parameter inference, large-scale analysis pipelines, and state-of-the-art visualization.
The minisymposium then turns to the engineering aspects that made these results possible: GPU acceleration, HPC integration, domain-specific code generation, interoperability with C++/Fortran/Python, and lessons from production deployments in research collaborations.
Finally, we highlight the organizational and community structures that sustain these efforts. We present embeddings of Julia in larger community initiatives like the HEP Software Foundation, the adoption of their software standards and training activities, which might converge into the building of a JuliaHEP ecosystem.
The session is designed for both scientists and technical practitioners seeking to understand why Julia works in physics, how it scales to real collaborations, and what barriers remain for large “C++/Python first” communities
Motivation and Goals
Most existing JuliaCon physics sessions focus either on scientific results or language features. Our minisymposium explicitly bridges both:
- Real science outcomes, not toy models.
- The enabling technical mechanisms, clearly articulated.
- The community practices that allow long-term sustainability.
Nuclear and particle physics provide particularly demanding use-cases: multi-scale descriptions, stochastic models, GPU kernels, large parameter inference pipelines, and tight incorporation of legacy C++/Fortran codebases (ROOT, Geant4, MC generators, etc.). Julia offers an attractive path toward performance and programmer ergonomics, but new users often encounter fragmentation, a lack of worked examples, or incomplete guidance for production deployments.
The minisymposium addresses these gaps in three complementary pillars:
Pillar A: Scientific Impact
Case studies of real research enabled by Julia: inference, experiment simulation, event generation, visualization, physics modeling, etc.Pillar B: Technical Insights
Engineering deep-dives: GPU kernel optimization, HPC workloads, performance portability, auto-differentiation, domain-specific code generation, interop with large C++/Python stacks.Pillar C: Community & Training
How training, governance, and shared standards foster maintainability and adoption: HSF involvement, training experiences, lessons learned in community building, and cross-disciplinary knowledge transfer.
Our goal is to produce actionable insights for current and upcoming users, while creating a realistic picture of adoption barriers and success stories within science-heavy environments.
Target Audience
- Researchers and engineers in computational science
- Developers working on HPC, GPU, and performance-critical applications
- Members of scientific collaborations (high-enregy, nuclear, astro, atomic and plasma physics)
- Physicists currently locked into Python/C++ toolchains
- Educators and community organizers
No domain expertise beyond general computational science is required.
Proposed Duration
180 min
Proposed Schedule
Opening (10 min)
Scope, motivation, framing, goals.
Block A: Scientific Impact (60 min)
Concrete physics results developed end-to-end in Julia.
Track A1: Astro/Particle Physics (20–25 min)
End-to-end simulation pipelines, inference, visualization, and event workflows.Track A2: Nuclear or Atomic Physics (20–25 min)
Solver pipelines, many–body modeling, large-scale integration, HPC workflows.
Depending on the submitted talks, this could also be a block of 5 lightning talks (10 minutes) distributed across tracks A1 and A2.
Break (10 min)
Coffee / bio / informal networking
Block B: Technical Insights (60 min)
Deep dives into how Julia enabled the science shown in Block A.
Track B1: GPU acceleration & performance portability (22–25 min)
Kernel optimization, KA.jl, memory layouts, scaling, portability, vectorization.Track B2: Interoperability & HPC ecosystem integration (22–25 min)
C++/Fortran/Python interoperability, wrapping large codebases, workflow integration.Audience Q&A (10–12 min)
Benchmarks, performance, adoption patterns, anti-patterns.
Again, depending on the submitted talks, this could also be a block of 5 lightning talks (10 minutes) distributed across tracks B1 and B2.
Break (10 min)
Coffee / bio / informal networking
Block C: Community & Training (30 min)
What sustains adoption and long-term contributions?
Track C1: Julia adoption & education in physics communities (20 min)
HSF Julia activities, ERUM-Data Julia workshops, teaching in C++/Python cultures, institutional hurdles, lessons learned.Audience Q&A (10 min)
Retention, onboarding, community success/failure stories.
Block D: Panel Discussion (30 min)
Moderated discussion with speakers.
Guiding questions:
- What works at scale? What still breaks?
- Interop vs rewriting: where is the boundary?
- What remains premature for physics workloads?
- How to onboard in legacy ecosystems (C++/Python)?
- What collaborations should exist, but don’t?
Goal:
Concrete actionable recommendations for tool builders and scientists.
Diversity and Inclusion
We will intentionally balance:
- gender and career levels (PhD -> senior)
- multiple subdomains (astro, nuclear, particle, atomic, plasma)
- institutional backgrounds (academia, labs)
- technical vs scientific perspectives
We aim to include early-career researchers who produced real Julia science, not only “project leaders”.
Why This Belongs at JuliaCon
- Demonstrates real-world scientific outcomes beyond toy demos
- Connects Julia’s differentiating technical features to practical performance
- Shows sustained community adoption beyond isolated success
- Builds bridges across physics subfields with common computational challenges
- Speaks to both new users and advanced practitioners
This minisymposium has direct educational impact and high visibility:
attendees leave with actionable directions and realistic expectations.
I am a particle physicist by training, currently working on the theoretical side of strong laser interactions and matter under extreme conditions at the Helmholtz-Zentrum Dresden-Rossendorf. For the past five years, I have conducted all my research in Julia, and I am the primary author of several Julia packages, including the QuantumElectrodynamics.jl framework, JuliaXRTS, and a maintainer at the JuliaHEP GitHub organization. In addition, for the past three years, I have served as one of the conveners of the JuliaHEP working group within the HEP Software Foundation, making me an active member of the Julia community in high-energy physics.