# Black hole spectral states revealed in GRMHD simulations with texture memory accelerated cooling

**Authors:** Pedro Naethe Motta, Jonatan Jacquemin-Ide, Rodrigo Nemmen, Matthew T. P. Liska, Alexander Tchekhovskoy  
**arXiv:** [2505.08855](https://arxiv.org/abs/2505.08855)  
**Date:** 2025-05-13  
**Summary page:** https://www.summarizepaper.com/en/paper/2505.08855v1/

## Key points

- Study conducted by Pedro Naethe Motta, Jonatan Jacquemin-Ide, Rodrigo Nemmen, Matthew T. P. Liska and Alexander Tchekhovskoy on X-ray binaries (XRBs) and spectral state changes
- Transitions in XRBs characterized by variations in hardness, luminosity, and accretion flow structure surrounding black holes
- Development of GPU-accelerated cooling toolkit for GRMHD simulations of accreting black holes
- Toolkit utilizes texture memory for quick retrieval of pre-computed values for accurate simulations
- Incorporation of radiation processes like bremsstrahlung, synchrotron emission, inverse Compton scattering, and Coulomb collisions in the toolkit
- Simulations focused on magnetically arrested disk scenario in low/hard states orbiting Kerr black holes with varying mass accretion rates
- Observation of geometrically thick two-temperature hot accretion flow at low accretion rates below ~0.01 Eddington rate
- Transition to cold single-temperature thin disk structure at higher mass accretion rates beyond threshold distance of ~50 gravitational radii (rg)
- Intricate patterns observed within thin disk region with intertwining filaments and hot thick flow environment
- GPU-accelerated cooling significantly faster than conventional radiation closure methods like M1 closure and storing lookup tables in global memory
- Study sheds light on interplay between mass accretion rates and resulting flow configurations in XRB systems through advanced GRMHD simulations

## Summary

Pedro Naethe Motta, Jonatan Jacquemin-Ide, Rodrigo Nemmen, Matthew T. P. Liska and Alexander Tchekhovskoy conducted a study on X-ray binaries (XRBs) and the changes in spectral states within these systems. These transitions are characterized by significant variations in hardness, luminosity and overall structure of the accretion flow surrounding black holes. To gain a better understanding of this phenomenon, the researchers developed a state-of-the-art GPU-accelerated cooling toolkit specifically designed for general relativistic magnetohydrodynamic (GRMHD) simulations of accreting black holes. This innovative toolkit utilizes texture memory to quickly retrieve pre-computed values necessary for accurate simulations. It incorporates various radiation processes such as bremsstrahlung, synchrotron emission, inverse Compton scattering and Coulomb collisions to provide a comprehensive representation of the physical interactions at play in XRB systems. By integrating this toolkit into a GRMHD code framework, the team conducted simulations focusing on a magnetically arrested disk scenario within the context of low/hard states in XRBs orbiting Kerr black holes. Through their simulations exploring mass accretion rates ranging from approximately $10^{-6}$ to 0.3 times the Eddington accretion rate ($\dot{M}_{\rm Edd}$), the researchers uncovered intriguing insights into the behavior of accretion flows at different rates. For low accretion rates below $\sim 0.01 \dot{M}_{\rm Edd}$, they observed the formation of a geometrically thick, low-density two-temperature hot accretion flow configuration. As the mass accretion rate increased beyond this threshold, the flow transitioned into a cold single-temperature thin disk structure at radial distances greater than approximately 50 gravitational radii ($r_g$). Remarkably, within this thin disk region, intricate patterns emerged as single-temperature thin filaments intertwined with a two-temperature hot thick flow environment. Notably, the team's GPU-accelerated cooling prescription utilizing texture memory proved to be significantly faster—approximately 3-5 times faster than conventional radiation closure methods like M1 closure and about 5 times faster than storing lookup tables in global memory. Overall, this study sheds light on the complex interplay between mass accretion rates and resulting accretion flow configurations in XRB systems through advanced GRMHD simulations enhanced by GPU-accelerated cooling techniques. The findings offer valuable insights into understanding black hole spectral states and pave the way for further advancements in astrophysical research related to high-energy phenomena in compact object environments.