FastPM: a new scheme for fast simulations of dark matter and halos
read the original abstract
We introduce FastPM, a highly-scalable approximated particle mesh N-body solver, which implements the particle mesh (PM) scheme enforcing correct linear displacement (1LPT) evolution via modified kick and drift factors. Employing a 2-dimensional domain decomposing scheme, FastPM scales extremely well with a very large number of CPUs. In contrast to COmoving-LAgrangian (COLA) approach, we do not require to split the force or track separately the 2LPT solution, reducing the code complexity and memory requirements. We compare FastPM with different number of steps ($N_s$) and force resolution factor ($B$) against 3 benchmarks: halo mass function from Friends of Friends halo finder, halo and dark matter power spectrum, and cross correlation coefficient (or stochasticity), relative to a high resolution TreePM simulation. We show that the modified time stepping scheme reduces the halo stochasticity when compared to COLA with the same number of steps and force resolution. While increasing $N_s$ and $B$ improves the transfer function and cross correlation coefficient, for many applications FastPM achieves sufficient accuracy at low $N_s$ and $B$. For example, $N_s=10$ and $B=2$ simulation provides a substantial saving (a factor of 10) of computing time relative to $N_s=40$, $B=3$ simulation, yet the halo benchmarks are very similar at $z=0$. We find that for abundance matched halos the stochasticity remains low even for $N_s=5$. FastPM compares well against less expensive schemes, being only 7 (4) times more expensive than 2LPT initial condition generator for $N_s=10$ ($N_s=5$). Some of the applications where FastPM can be useful are generating a large number of mocks, producing non-linear statistics where one varies a large number of nuisance or cosmological parameters, or serving as part of an initial conditions solver.
This paper has not been read by Pith yet.
Forward citations
Cited by 9 Pith papers
-
FAIR Universe Weak Lensing ML Uncertainty Challenge: Handling Uncertainties and Distribution Shifts for Precision Cosmology
A standardized weak lensing benchmark dataset with realistic systematics is released alongside a two-phase ML uncertainty challenge to advance data-efficient and robust cosmological analysis.
-
Perturbation-theory informed integrators for cosmological simulations
LPT-matched integrators for cosmological simulations outperform FastPM with O(1-100) timesteps while convergence is limited to order 3/2 post-shell-crossing due to acceleration field irregularity.
-
Towards Practical Field-Level Inference for Weak Lensing
Field-level inference from weak lensing maps yields significantly tighter cosmological constraints than power-spectrum analysis when using the same forward-modeling pipeline, especially on small scales.
-
Learning the Universe: Posterior Reliability of Neural Generative Models in High-Dimensional Field-Level Inference of Cosmic Initial Conditions
Generative models for cosmological field-level inference can reproduce posterior means and cross-correlations yet fail to capture correct uncertainty geometry when validated against HMC reference samples.
-
Cosmic variance or galaxy bias? Disentangling finite-volume and galaxy formation effects in cosmological analysis
Develops a galaxy-biasing formalism for cosmic variance using perturbation theory and tests it on the non-linear BAO shift against N-body simulations.
-
Cosmological Analysis with Calibrated Neural Quantile Estimation and Approximate Simulators
Calibrated NQE enables unbiased field-level cosmological inference from 2D density maps by training mostly on approximate PM simulations and calibrating with ~100 PP simulations.
-
$\textit{BlockFormer}$ : Transformer-based inference from interaction maps
BlockFormer applies a transformer architecture to localize centromeres from Hi-C maps by training on abundant synthetic interaction data generated by a custom simulator.
-
$\textit{BlockFormer}$ : Transformer-based inference from interaction maps
A transformer architecture trained on simulated interaction maps recovers centromere genomic positions from real Hi-C data across species with varying genome sizes.
-
Large-scale structures of the Universe: physics, phenomenology, statistics
Lecture series on the physics, phenomenology, and statistics of large-scale cosmic structure evolution and non-Gaussian predictions.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.