SciPost Submission Page
Bonsai-SPH: A GPU accelerated astrophysical Smoothed Particle Hydrodynamics code
by Jeroen Bédorf, Simon Portegies Zwart
This is not the latest submitted version.
This Submission thread is now published as
Submission summary
| Authors (as registered SciPost users): | Jeroen Bédorf · Simon Portegies Zwart |
| Submission information | |
|---|---|
| Preprint Link: | https://arxiv.org/abs/1909.07439v1 (pdf) |
| Date submitted: | 2019-09-20 02:00 |
| Submitted by: | Bédorf, Jeroen |
| Submitted to: | SciPost Physics |
| Ontological classification | |
|---|---|
| Academic field: | Astronomy |
| Specialties: |
|
| Approach: | Computational |
Abstract
We present the smoothed-particle hydrodynamics simulation code, Bonsai-SPH, which is a continuation of our previously developed gravity-only hierarchical $N$-body code (called Bonsai). The code is optimized for Graphics Processing Unit (GPU) accelerators which enables researchers to take advantage of these powerful computational resources. Bonsa-SPH produces simulation results comparable with state-of-the-art, CPU based, codes, but using an order of magnitude less computation time. The code is freely available online and the details are described in this work.
Current status:
Reports on this Submission
Report #1 by Mark Vogelsberger (Referee 1) on 2020-1-1 (Invited Report)
- Cite as: Mark Vogelsberger, Report on arXiv:1909.07439v1, delivered 2020-01-01, doi: 10.21468/SciPost.Report.1425
Strengths
1 - The paper presents a very clear description of the SPH implementation in Bonsai-SPH.
2 - The methods and test problems are clearly laid out.
3 - The figures are very polished.
Weaknesses
1 - It would be nice to see some small scientific application problems in the paper beyond the test problems. Maybe the authors can add something along these lines?
2 - The paper does not detail how the code performs on parallel problems. Can the authors add a figure demonstrating well parallel processing works on multiple GPUs?
3 - The last years have seen various novel SPH implementations. It would be good if the introduction could discuss some of those. The authors should also discuss a bit more in detail the advantages and disadvantages of SPH compared to other techniques.
Report
The paper "Bonsai-SPH: A GPU accelerated astrophysical Smoothed
Particle Hydrodynamics code" by Bedorf and Zwart presents the smoothed-particle hydrodynamics simulation code, Bonsai-SPH,
which is a continuation of the gravity-only hierarchical N-body code Bonsai. The code is optimized for GPU accelerators. The authors demonstrate that Bonsai-SPH produces simulation results comparable
with state-of-the-art, CPU based, codes, but using an order of magnitude less
computation time.
Overall this paper presents a concise and clear description of the methods and test problems. There are a few points where the authors could expand the paper slightly. I discuss those under 'weaknesses'. It would be nice to see some of these suggestions being implemented by the authors. However, this is not a requirement since the paper is in its current form already publishable.
Requested changes
1 - Add some small scientific application at the end that goes beyond the pure test problems.
2 - Discuss in more depth the parallel aspects of the code.
3 - Extend the intro by a more detailed review of recent efforts in SPH implementations and also discuss in more detail the advantages and disadvantages of SPH.