arXiv subjects

Hannes Vogt

Publications and source records attributed to Hannes Vogt.

10 recordsLinked to original sources

arXiv2026

Integrating a Python Dynamical core into ICON

The transition of Earth-system models to exascale is often hindered by rigid, monolithic Fortran codebases and maintenance-heavy compiler directives. While high-level DSLs offer a solution, they frequently fail due to cumbersome integration. We present the integration of a Python-based ICON dynamical core into the original Fortran simulation code. Leveraging the GT4Py DSL and the Data-Centric (DaCe) optimization framework, we demonstrate that high-level Python can be seamlessly integrated into legacy infrastructure without performance loss. Our results challenge the assumption that Python orchestration introduces prohibitive HPC overhead. In production-grade global simulations, our Python dynamical core achieves a 20--30\% performance improvement over the highly-optimized Fortran+OpenACC implementation, with a 10\% improvement on the total time for a coupled setup. Driven by advanced data-flow optimizations and automated kernel fusion, this approach replaces hardware-entangled directives by generating optimized device code from a single, portable Python source. This work proves that Python can provide a sustainable, efficient, and hardware-agnostic future for global climate modeling.

cs.DC↗

arXiv2025

$O(a)$-improved QCD+QED Wilson Dirac operator on GPUs

Markov Chain Monte Carlo simulations of lattice Quantum Chromodynamics (QCD) are the only known tool to investigate non-perturbatively the theory of the strong interaction and are required to perform precision tests of the Standard Model of Particle Physics. As the Markov Chain is a serial process, the sole option for improving the sampling rate is accelerating each individual update step. Heterogeneous clusters of GPU-accelerated nodes offer large total memory bandwidth which can be used to speed-up our application, openQxD-1.1, which is dominated by inversions of the Dirac operator, a large sparse matrix. In this work we investigate offloading the inversion to GPU using the lattice-QCD library QUDA, and our early results demonstrate a significant potential speed-up in the time-to-solution for state-of-the-art problem sizes. Minimal extensions to the existing QUDA library are required for our specific physics programme while greatly enhancing the performance portability of our code and retaining the reliability and robustness of existing applications in openQxD-1.1. Our new interface will enable us to utilize pre-exascale infrastructure and reduce the systematic uncertainty in our physics predictions by incorporating the effects of quantum electromagnetism (QED) in our simulations.

hep-lat↗

arXiv2023

GT4Py: High Performance Stencils for Weather and Climate Applications using Python

All major weather and climate applications are currently developed using languages such as Fortran or C++. This is typical in the domain of high performance computing (HPC), where efficient execution is an important concern. Unfortunately, this approach leads to implementations that intermix optimizations for specific hardware architectures with the high-level numerical methods that are typical for the domain. This leads to code that is verbose, difficult to extend and maintain, and difficult to port to different hardware architectures. Here, we propose a different strategy based on GT4Py (GridTools for Python). GT4Py is a Python framework to write weather and climate applications that includes a high-level embedded domain specific language (DSL) to write stencil computations. The toolchain integrated in GT4Py enables automatic code-generation,to obtain the performance of state-of-the-art C++ and CUDA implementations. The separation of concerns between the mathematical definitions and the actual implementations allows for performance portability of the computations on a wide range of computing architectures, while being embedded in Python allows easy access to the tools of the Python ecosystem to enhance the productivity of the scientists and facilitate integration in complex workflows. Here, the initial release of GT4Py is described, providing an overview of the current state of the framework and performance results showing how GT4Py can outperform pure Python implementations by orders of magnitude.

cs.DC↗

arXiv2016

Gribov horizon and Gribov copies effect in lattice Coulomb gauge

Following a recent proposal by Cooper and Zwanziger we investigate via $SU(2)$ lattice simulations the effect on the Coulomb gauge propagators and on the Gribov-Zwanziger confinement mechanism of selecting the Gribov copy with the smallest non-trivial eigenvalue of the Faddeev-Popov operator, i.e.~the one closest to the Gribov horizon. Although such choice of gauge drives the ghost propagator towards the prediction of continuum calculations, we find that it actually overshoots the goal. With increasing computer time, we observe that Gribov copies with arbitrarily small eigenvalues can be found. For such a method to work one would therefore need further restrictions on the gauge condition to isolate the physically relevant copies, since e.g.~the Coulomb potential $V_C$ defined through the Faddeev-Popov operator becomes otherwise physically meaningless. Interestingly, the Coulomb potential alternatively defined through temporal link correlators is only marginally affected by the smallness of the eigenvalues.

hep-lat↗

arXiv2015

Lattice QCD Green's functions in maximally Abelian gauge: infrared Abelian dominance and the quark sector

On lattice gauge field configurations with 2+1 dynamical quark flavors, we investigate the momentum space quark and gluon propagators in the combined maximally Abelian plus $U(1)_3\times U(1)_8$ Landau gauge. We extract the gluon fields from the lattice link variables and study the diagonal and off-diagonal gluon propagators. We find that the infrared region of the transverse diagonal gluon propagator is strongly enhanced compared to the off-diagonal propagator. The Dirac operator from the Asqtad action is inverted on the diagonal and off-diagonal gluon backgrounds separately. In agreement with the hypothesis of infrared Abelian dominance, we find that the off-diagonal gluon background hardly gives rise to any non-trivial quark dynamics while the quark propagator from the diagonal gluon background closely resembles its Landau gauge counterpart.

hep-lat↗

arXiv2014

cuLGT: Lattice Gauge Fixing on GPUs

We adopt CUDA-capable Graphic Processing Units (GPUs) for Landau, Coulomb and maximally Abelian gauge fixing in 3+1 dimensional SU(3) and SU(2) lattice gauge field theories. A combination of simulated annealing and overrelaxation is used to aim for the global maximum of the gauge functional. We use a fine grained degree of parallelism to achieve the maximum performance: instead of the common 1 thread per site strategy we use 4 or 8 threads per lattice site. Here, we report on an improved version of our publicly available code (www.cuLGT.com and github.com/culgt) which again increases performance and is much easier to include in existing code. On the GeForce GTX 580 we achieve up to 470 GFlops (utilizing 80% of the theoretical peak bandwidth) for the Landau overrelaxation code.

hep-lat↗

arXiv2013

Coulomb gauge on the lattice: From zero to finite temperature

Our previous studies of Coulomb gauge Yang-Mills theory are extended to finite temperature. We investigate the SU(2) static gluon and ghost propagators and show results for the Coulomb potential, with a focus on the Gribov ambiguity. To compute these quantities at high temperatures and to solve scaling violations we use the anisotropic Wilson gauge action.

hep-lat↗

arXiv2013

Gauge fixing in lattice QCD with multi-GPUs

Here we present the cuLGT code for gauge fixing in lattice gauge field theories with graphic processing units (GPUs). Implementations for SU(3) Coulomb, Landau and maximally Abelian gauge fixing are available and the overrelaxation, stochastic relaxation and simulated annealing algorithms are supported. Performance results for single and multi-GPUs are given.

hep-lat↗

arXiv2013

Coulomb, Landau and Maximally Abelian Gauge Fixing in Lattice QCD with Multi-GPUs

A lattice gauge theory framework for simulations on graphic processing units (GPUs) using NVIDIA's CUDA is presented. The code comprises template classes that take care of an optimal data pattern to ensure coalesced reading from device memory to achieve maximum performance. In this work we concentrate on applications for lattice gauge fixing in 3+1 dimensional SU(3) lattice gauge field theories. We employ the overrelaxation, stochastic relaxation and simulated annealing algorithms which are perfectly suited to be accelerated by highly parallel architectures like GPUs. The applications support the Coulomb, Landau and maximally Abelian gauges. Moreover, we explore the evolution of the numerical accuracy of the SU(3) valued degrees of freedom over the runtime of the algorithms in single (SP) and double precision (DP). Therefrom we draw conclusions on the reliability of SP and DP simulations and suggest a mixed precision scheme that performs the critical parts of the algorithm in full DP while retaining 80-90% of the SP performance. Finally, multi-GPUs are adopted to overcome the memory constraint of single GPUs. A communicator class which hides the MPI data exchange at the boundaries of the lattice domains, via the low bandwidth PCI-Bus, effectively behind calculations in the inner part of the domain is presented. Linear scaling using 16 NVIDIA Tesla C2070 devices and a maximum performance of 3.5 Teraflops on lattices of size down to 64^3 x 256 is demonstrated.

hep-lat↗

arXiv2012

Gauge fixing using overrelaxation and simulated annealing on GPUs

We adopt CUDA-capable Graphic Processing Units (GPUs) for Coulomb, Landau and maximally Abelian gauge fixing in 3+1 dimensional SU(3) lattice gauge field theories. The local overrelaxation algorithm is perfectly suited for highly parallel architectures. Simulated annealing preconditioning strongly increases the probability to reach the global maximum of the gauge functional. We give performance results for single and double precision. To obtain our maximum performance of ~300 GFlops on NVIDIA's GTX 580 a very fine grained degree of parallelism is required due to the register limits of NVIDIA's Fermi GPUs: we use eight threads per lattice site, i.e., one thread per SU(3) matrix that is involved in the computation of a site update.

hep-lat↗