HARD: A Performance Portable Radiation Hydrodynamics Code based on FleCSI Framework

HARD is an open-source, performance-portable radiation hydrodynamics code built on the FleCSI framework and Kokkos that enables efficient simulations across diverse hardware architectures while ensuring scientific reliability through automated regression testing and community-driven development.

Original authors: Julien Loiseau, Hyun Lim, Andrés Yagüe López, Mammadbaghir Baghirzade, Shihab Shahriar Khan, Yoonsoo Kim, Sudarshan Neopane, Alexander Strack, Farhana Taiyebah, Benjamin K. Bergen

Published 2026-05-19
📖 4 min read☕ Coffee break read

Original authors: Julien Loiseau, Hyun Lim, Andrés Yagüe López, Mammadbaghir Baghirzade, Shihab Shahriar Khan, Yoonsoo Kim, Sudarshan Neopane, Alexander Strack, Farhana Taiyebah, Benjamin K. Bergen

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to simulate a massive, chaotic event—like a star exploding or a fusion bomb detonating. To do this, you need a computer program that can track two things happening at once: how gas moves (hydrodynamics) and how light energy (radiation) heats and pushes that gas. This is called Radiation Hydrodynamics.

The paper introduces a new software tool called HARD (Hydrodynamics And Radiation Diffusion) designed to solve these complex puzzles. Here is how it works, explained through simple analogies:

1. The "Universal Adapter" (Performance Portability)

Think of the world's supercomputers like different types of vehicles: some are sedans (laptops), some are trucks (standard clusters), and some are massive, custom-built race cars (the world's most powerful supercomputers with both CPUs and GPUs).

Usually, software is like a car engine built for only one specific vehicle. If you want to run it on a different car, you have to rebuild the engine from scratch. HARD is different. It is built on a "universal adapter" called FleCSI.

  • The Analogy: Imagine a video game controller that automatically reconfigures itself to work with a PlayStation, an Xbox, or a PC without you changing the buttons. HARD does this for computers. It writes the physics code once, and then it automatically translates that code to run efficiently on anything from a laptop to a giant supercomputer, whether that machine uses standard processors or specialized graphics cards (GPUs).

2. The "Task Manager" (Orchestration)

Simulating a star explosion involves millions of tiny calculations happening at the same time.

  • The Analogy: Imagine a construction site. Instead of one foreman telling every single worker what to do one by one (which is slow), HARD acts like a smart project manager. It breaks the job into small "tasks" (like "pour concrete here" or "measure this beam") and hands them out to a team of workers.
  • The Magic: If the workers are in one building, the manager uses one communication style (MPI). If they are in different buildings, it uses another (Legion or HPX). The workers (the physics calculations) don't need to know how they are being managed; they just do their job. This allows the software to scale up or down instantly.

3. The "Double-Check System" (Verification)

In science, you can't just trust the numbers; you have to prove they are right.

  • The Analogy: HARD comes with a built-in "training manual" and a "pop quiz." It automatically runs famous, well-known test problems (like the "Sod shock tube," which is like a standard physics exam question everyone knows the answer to).
  • The Result: The software compares its own answer to the known "correct" answer. If they match, the software passes the test. This ensures that when scientists use it for new, unknown problems, the results are trustworthy.

4. What Does It Actually Simulate?

The paper shows HARD working on a few specific scenarios:

  • The Shock Tube: Like a dam breaking, where high-pressure gas rushes into low-pressure gas, creating a shockwave. HARD predicted the wave's shape perfectly.
  • Heating and Cooling: Imagine a pot of water sitting next to a heater. The paper shows HARD accurately calculating how the water heats up until it matches the heater's temperature, and how it cools down if the heater is turned off.
  • The "Radiation Kick": In some scenarios, light energy is so strong it creates its own shockwaves. HARD showed that when you add radiation, these shockwaves form faster and behave differently than when you only look at the gas.
  • The "Swirling Fluid" (Kelvin-Helmholtz Instability): Imagine two rivers flowing past each other at different speeds, creating a swirling, messy boundary. The paper found that adding radiation to this mix makes the swirls grow and become chaotic much faster than without it.

5. Speed and Scale

The authors tested HARD on a massive supercomputer called Chicoma.

  • The Analogy: They tried to solve a puzzle by adding more and more people to the team.
  • The Result: As they added more computer "workers" (nodes), the speed of the simulation increased almost perfectly. It didn't slow down due to communication delays.
  • The GPU Boost: When they tested it on computers with powerful graphics cards (GPUs), a single graphics card was 7 times faster than a standard computer processor.

Summary

HARD is a new, open-source tool for scientists to simulate how matter and light interact in extreme environments. Its main superpower is that it is portable (runs on any computer), reliable (proves its own math is right), and fast (scales up to the biggest supercomputers). It is designed to help researchers understand everything from how stars explode to how we might create clean fusion energy.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →