arXiv subjects

Lubomir Riha

Publications and source records attributed to Lubomir Riha.

8 recordsLinked to original sources

arXiv2026

Accelerating cosmological simulations on GPUs: a step towards sustainability and green-awareness

The increasing complexity and scale of cosmological N-body simulations, driven by astronomical surveys like Euclid, call for a paradigm shift towards more sustainable and energy-efficient high-performance computing (HPC). The rising energy consumption of supercomputing facilities poses a significant environmental and financial challenge. In this work, we build upon a recently developed GPU implementation of pinocchio, a widely-used tool for the fast generation of dark matter (DM) halo catalogues, to investigate energy consumption. Using a different resource configuration, we confirmed the time-to-solution behavior observed in a companion study, and we use these runs to compare time-to-solution with energy-to-solution. By profiling the code on various HPC platforms with a newly developed implementation of the Power Measurement Toolkit (PMT), we demonstrate an 8x reduction in energy-to-solution and 8x speed-up in time-to-solution compared to the CPU-only version. Taken together, these gains translate into an overall efficiency improvement of up to 64x. Our results show that the GPU-accelerated pinocchio not only achieves substantial speed-up, making the generation of large-scale mock catalogues more tractable, but also significantly reduces the energy footprint of the simulations. This work represents an step towards ``green-aware" scientific computing in cosmology, proving that performance and sustainability can be simultaneously achieved.

astro-ph.IM↗

arXiv2025

EuroHPC SPACE CoE: Redesigning Scalable Parallel Astrophysical Codes for Exascale

High Performance Computing (HPC) based simulations are crucial in Astrophysics and Cosmology (A&C), helping scientists investigate and understand complex astrophysical phenomena. Taking advantage of exascale computing capabilities is essential for these efforts. However, the unprecedented architectural complexity of exascale systems impacts legacy codes. The SPACE Centre of Excellence (CoE) aims to re-engineer key astrophysical codes to tackle new computational challenges by adopting innovative programming paradigms and software (SW) solutions. SPACE brings together scientists, code developers, HPC experts, hardware (HW) manufacturers, and SW developers. This collaboration enhances exascale A&C applications, promoting the use of exascale and post-exascale computing capabilities. Additionally, SPACE addresses high-performance data analysis for the massive data outputs from exascale simulations and modern observations, using machine learning (ML) and visualisation tools. The project facilitates application deployment across platforms by focusing on code repositories and data sharing, integrating European astrophysical communities around exascale computing with standardised SW and data protocols.

astro-ph.IM↗

arXiv2025

The PLUTO Code on GPUs: A First Look at Eulerian MHD Methods

We present preliminary performance results of gPLUTO, the new GPU-optimized implementation of the PLUTO code for computational plasma astrophysics. Like its predecessor, gPLUTO employs a finite-volume formulation to numerically solve the equations of magnetohydrodynamics (MHD) in multiple spatial dimensions. Still, this new implementation is a complete rewrite in C++ and leverages the OpenACC programming model to achieve acceleration on NVIDIA GPUs. While a more comprehensive description of the code and its several other modules will be presented in a future paper, here we focus on some preparatory results that demonstrate the code potential and performance on pre exa-scale parallel architectures.

physics.plasm-ph↗

arXiv2025

Towards Exascale Computing for Astrophysical Simulation Leveraging the Leonardo EuroHPC System

Developing and redesigning astrophysical, cosmological, and space plasma numerical codes for existing and next-generation accelerators is critical for enabling large-scale simulations. To address these challenges, the SPACE Center of Excellence (SPACE-CoE) fosters collaboration between scientists, code developers, and high-performance computing experts to optimize applications for the exascale era. This paper presents our strategy and initial results on the Leonardo system at CINECA for three flagship codes, namely gPLUTO, OpenGadget3 and iPIC3D, using profiling tools to analyze performance on single and multiple nodes. Preliminary tests show all three codes scale efficiently, reaching 80% scalability up to 1,024 GPUs.

cs.DC↗

arXiv2025

Heterogeneous Memory Pool Tuning

We present a lightweight tool for the analysis and tuning of application data placement in systems with heterogeneous memory pools. The tool allows non-intrusively identifying, analyzing, and controlling the placement of individual allocations of the application. We use the tool to analyze a set of benchmarks running on the Intel Sapphire Rapids platform with both HBM and DDR memory. The paper also contains an analysis of the performance of both memory subsystems in terms of read/write bandwidth and latency. The key part of the analysis is to focus on performance if both subsystems are used together. We show that only about 60% to 75% of the data must be placed in HBM memory to achieve 90% of the potential performance of the platform on those benchmarks.

cs.PF↗

arXiv2025

Methodology for GPU Frequency Switching Latency Measurement

The development of exascale and post-exascale HPC and AI systems integrates thousands of CPUs and specialized accelerators, making energy optimization critical as power costs rival hardware expenses. To reduce consumption, frequency and voltage scaling techniques are widely used, but their effectiveness depends on adapting to application demands in real-time. However, frequency scaling incurs a switching latency, impacting the responsiveness of dynamic tuning approaches. We propose a methodology to systematically evaluate the frequency switching latency of accelerators, with an implementation for CUDA. Our approach employs an artificial iterative workload designed for precise runtime measurements at different frequencies. The methodology consists of three phases: (1) measuring workload execution time across target frequencies, (2) determining switching latency by tracking the transition from an initial to a target frequency, and (3) filtering out outliers caused by external factors such as CUDA driver management or CPU interruptions. A robust statistical system ensures accuracy while minimizing execution time. We demonstrate this methodology on three Nvidia GPUs - GH200, A100, and RTX Quadro 6000 - revealing significant variations in switching latency. These findings are crucial for designing energy-efficient runtime systems, helping determine optimal frequency change rates and avoiding transitions with excessive overhead.

cs.DC↗

arXiv2023

Scalable Flow Simulations with the Lattice Boltzmann Method

The primary goal of the EuroHPC JU project SCALABLE is to develop an industrial Lattice Boltzmann Method (LBM)-based computational fluid dynamics (CFD) solver capable of exploiting current and future extreme scale architectures, expanding current capabilities of existing industrial LBM solvers by at least two orders of magnitude in terms of processor cores and lattice cells, while preserving its accessibility from both the end-user and software developer's point of view. This is accomplished by transferring technology and knowledge between an academic code (waLBerla) and an industrial code (LaBS). This paper briefly introduces the characteristics and main features of both software packages involved in the process. We also highlight some of the performance achievements in scales of up to tens of thousand of cores presented on one academic and one industrial benchmark case.

physics.comp-ph↗

arXiv2020

Toward an End-to-End Auto-tuning Framework in HPC PowerStack

Efficiently utilizing procured power and optimizing performance of scientific applications under power and energy constraints are challenging. The HPC PowerStack defines a software stack to manage power and energy of high-performance computing systems and standardizes the interfaces between different components of the stack. This survey paper presents the findings of a working group focused on the end-to-end tuning of the PowerStack. First, we provide a background on the PowerStack layer-specific tuning efforts in terms of their high-level objectives, the constraints and optimization goals, layer-specific telemetry, and control parameters, and we list the existing software solutions that address those challenges. Second, we propose the PowerStack end-to-end auto-tuning framework, identify the opportunities in co-tuning different layers in the PowerStack, and present specific use cases and solutions. Third, we discuss the research opportunities and challenges for collective auto-tuning of two or more management layers (or domains) in the PowerStack. This paper takes the first steps in identifying and aggregating the important R&D challenges in streamlining the optimization efforts across the layers of the PowerStack.

cs.PF↗