FUGUE: A New GPU-accelerated General Relativistic Magnetohydrodynamics Code with Weak-Field Multipole Self-Gravity and Quasi-Stationary Kerr Black Hole Evolution

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: I'm Vera.

Jocelyn: And I'm Jocelyn, here with Subrahmanyan.

Vera: We're looking at the paper: "FUGUE: A New GPU-accelerated General Relativistic Magnetohydrodynamics Code with Weak-Field Multipole Self-Gravity and Quasi-Stationary Kerr Black Hole Evolution".

Jocelyn: The gist The code FUGUE presents a new GPU-accelerated General Relativistic Magnetohydrodynamics (GRMHD) code that supports optional quasi-stationary Kerr black hole evolution and weak-field multipole self-gravity,

Vera: First, who's behind it and why it matters.

Title and authors: Vera: So, the title itself tells you a lot about what this code does. It’s about FUGUE being GPU-accelerated for general relativistic magnetohydrodynamics, and it specifically handles two things: a quasi-stationary update of the Kerr black hole mass and spin, and weak-field multipole self-gravity.

Jocelyn: That sounds like a lot of moving parts. They’re not just simulating the fluid dynamics around a fixed black hole; they’re letting the black hole itself evolve while also adding this correction for how the fluid's own gravity affects things, using a method that's good for GPUs.

Subrahmanyan: The GPU acceleration part is important because these simulations are computationally intensive, and if you can run them faster, you can do more complex physics. It suggests they’ve managed to make this kind of simulation practical for studying those extreme environments we see in collapsars.

The paper's summary: Vera: The summary explains that FUGUE is a new GPU-accelerated code designed to solve the covariant equations of ideal magnetohydrodynamics in general relativity. It includes the continuity equation, the conservation of energy–momentum, and the induction equation.

Jocelyn: That's saying it handles all the fundamental physics of how magnetic fields move and how matter flows within curved spacetime, which is a huge step for any GRMHD simulation.

Subrahmanyan: And they mention this optional quasi-stationary update of the Kerr black hole mass and spin happens by calculating fluxes through a radial inner boundary located inside the event horizon. That’s a specific way to handle that evolution.

The paper's improvements: Vera: One big improvement mentioned is their method for handling self-gravity, they use a weak-field module computed with a spherical harmonic Poisson solver. They treat this self-gravity as a scalar potential that generates a four-force on top of the Kerr background metric instead of changing the metric itself.

Jocelyn: So it’s not solving Einstein's equations perfectly for the gravity, but it’s an effective approximation that treats self-gravity as a force acting on the existing spacetime. That makes sense given how hard those full solutions are to get.

Subrahmanyan: They define the effective gravitational density used for that Poisson equation as rho equals E plus S, where E is the energy density measured by an Eulerian observer and S is the trace of the spatial stress tensor. The source term for this Poisson equation includes things like internal energy, pressure, and magnetic field energy.

Conclusion: Vera: To wrap up on FUGUE, they validated this code with hydrodynamic tests, magnetohydrodynamic tests, general relativistic tests, and convergence tests for the self-gravity module to show it’s working accurately.

Jocelyn: And performance-wise, they report that on a single NVIDIA GH200 GPU with linear reconstruction, the code achieved more than four hundred seventy million zone-cycles per second. That’s a lot of work done in one go.

Subrahmanyan: The application they show is for simulating a self-gravitating, Toomre-unstable, magnetized accretion disk around a Kerr black hole with a mass of one solar mass <ref:2610.20869#pg1,a self-gravitating, Toomre-unstable, magnetized accretion disk>. They show that in this model, the total normalized Fourier-mode power P(t) is about a hundred times larger than in the model without self-gravity.

Vera: So what does this mean for us listening? It means we can now study those gravitational instabilities in accretion disks with a level of detail that was previously out of reach because of the computational cost.

Jocelyn: It opens up a way to track how gravitational instabilities affect things like mass accretion rates and the Blandford–Znajek luminosity, which might contribute to gamma-ray burst variability.

Subrahmanyan: And they point out their limitation: they use an effective weak-field approximation for multipole self-gravity, and near the black hole where self-gravity is subdominant compared to the geometric source term, errors in that force should only have a small local effect.

Vera: They also mentioned future work plans, like implementing microphysics and neutrino transport methods such as the M1 method for neutrino transport.

Jocelyn: It’s a really high-performance code for tracking spin equilibrium dynamically and very efficiently, which is important for studying black hole spin-down in collapsar simulations.

Subrahmanyan: Ultimately, this FUGUE paper gives researchers a highly optimized code for solving GRMHD equations and demonstrates that during the phase when the disk is gravitationally unstable, the mass accretion rate is more variable and it also affects the Bland–Znajek luminosity.