Map to venue, and event information
Draft Agenda
Day 1: Thursday 19th March 2026
- 09:30 - 10:00 Day 1 Register and Welcome Coffee
- 10:00 - 10:10 Welcome from EPCC
- 10:10 - 11:25 Talk Session 1
- 10:10 - 10:35 Alexander Morozov (University of Edinburgh)
- 10:35 - 11:00 Xin Zhou (University of Leeds)
- 11:00 - 11:25 Caitlin McAuley (EPSRC-UKRI)
- 11:25 - 11:45 Coffee Break
- 11:45 - 12:35 Talk Session 2
- 11:45 - 12:10 Marc A. Little (Heriot-Watt University)
- 12:10 - 12:35 Nunzio Palumbo (Rolls-Royce)
- 12:35 - 13:30 Lunch
- 13:30 - 14:45 Interactive session
- 14:45 - 15:00 Coffee break
- 15:00 - 16:15 Talk Session 3 (GPU-eCSE talks)
- 15:00 - 15:15 Paul Bartholomew (EPCC, University of Edinburgh)
- 15:15 - 15:30 TBC (University of Manchester)
- 15:30 - 15:45 Paddys Roddy (UCL)
- 15:45 - 16:00 Ian Bush (STFC)
- 16:00 - 16:15 Nick Brown (EPCC, University of Edinburgh)
- 16:15 - 16:30 Coffee break
- 16:30 - 18:00 Lightning Talk Session
- 18:00 - 19:30 Poster Session / Drink Reception
- 19:30 Day 1 Finish
Day 2: Friday 20th March 2026
- 08:30 - 09:00 Day 2 Arrival Coffee
- 09:00 - 10:00 User Forum (Hybrid Session)
- 10:00 - 10:10 Mini Break
- 10:10 - 11:25 Talk Session 4
- 10:10 - 10:35 James Panton (University of Cologne) - Remote
- 10:35 - 11:00 Gabriele C. Sosso (University of Warwick)
- 11:00 - 11:25 Niamh O’Neill (Max Planck Institute for Polymer Research)
- 11:25 - 11:40 Coffee Break
- 11:40 - 12:10 Prof. Mark Parsons (EPCC, University of Edinburgh) - Keynote
- 12:10 - 13:00 Panel Session
- 13:00 - 14:00 Lunch
- 14:00 - 15:45 NetworkPlus Session
- 15:45 - 16:00 ARCHER2 Celebration of Science 2026 wrap-up
- 16:00 Day 2 Finish
Confirmed speakers and talk titles
Lightning Talks
Lightning Talk Abstracts
1 Molecular Dynamics Investigation of Crosslink Density Effects in Poly (vinyl alcohol) Hydrogel
Arun Srikanth Sridhar, Heriot Watt University
Crosslink density is a key structural parameter governing the physical behavior of polymer hydrogels, yet its molecular-level influence on network structure and dynamics remains incompletely understood. In this ongoing study, atomistic molecular dynamics simulations are employed to investigate the effects of crosslink density in poly(vinyl alcohol) (PVA) hydrogels.
Network models are constructed across a controlled range of crosslink densities while preserving chemical composition and stoichiometry. Simulations are performed in the anhydrous state to isolate intrinsic network behavior independent of solvent-mediated effects. The objective is to systematically quantify how increasing constraint density alters local packing, segmental mobility, free volume characteristics, and thermophysical response.
Structural analysis will include radial distribution functions and hydrogen-bond characterization to assess changes in intermolecular organization with increasing crosslink concentration. Dynamic properties will be evaluated using mean-squared displacement and time-correlation functions to examine the suppression of segmental relaxation as network constraints increase. Temperature-dependent simulations are being conducted to assess trends in glass transition behavior, and equilibrium Green–Kubo methods are planned to explore potential changes in thermal conductivity associated with enhanced connectivity.
By establishing a consistent computational framework for varying crosslink density, this work aims to provide molecular-level insight into structure–dynamics relationships in PVA hydrogels and to clarify how network constraint density governs polymer relaxation and thermophysical behavior.
2 New, high-performance algorithms for the computation of the Hardy Function
David Lewis, University of Liverpool
The Riemann zeta-function ζ(s) for s ∈ ℂ, is a complex function fundamental to prime number theory. Computation of ζ(s) along the critical line s=1⁄2+It, t ∈ ℝ, is important for the distribution of primes, since all the zeta-zeros are predicted to lie here (the Riemann Hypothesis). Hardy’s Z-function Z(t)=eIθ(t) ζ(1⁄2+It), θ(t) ∈ ℝ, defines the amplitude of the zeta-function along the critical line. Such computations also provide a means of Riemann Hypothesis verification, within closed intervals, and estimates on the bounds of ζ(1⁄2+It), central to the Lindelöf Hypothesis.
Until recently, the most efficient means of computing Z(t) employed the Riemann-Siegel (RS) formula, an O(√t) operational method. In 2011, this was superseded by an algorithm devised by G. A. Hiary, requiring just O(t1/3){log(t)}κ ) operations. The methodology involved the sub-division of the RS formula into sequences of quadratic Gauss sums of length 𝑁. Such sums are amenable to rapid computation, in 𝑂(𝑙𝑜𝑔(𝑁)) operations, using a recursive scheme. This poster will present an advancement on these ideas. Central to this work is a new asymptotic formula for 𝑍(𝑡), with some interesting analytical properties. Computationally, this new formulation allows one to improve on the quadratic sum expansion and express 𝑍(𝑡) in terms of sub-sequences of mth-order Gauss sums. In the cubic m=3 case, these sums can be computed rapidly, utilising a similar scheme to that used for the quadratics. The result is a more efficient algorithm, requiring only O((t ⁄ εt )1/4 {log(t)}3 ) operations. Sample computations, using the open-source code developed for this eCSE Project and available on GitHub repository, support these findings.
3 Irradiation tolerance of nanocrystalline Fe and FeCr: Large scale atomistic simulations and matched experiments
Ivan Tolkachev, University of Oxford
During operation, deuterium-tritium fusion devices generate high-energy neutrons that penetrate the lattice of reactor components, triggering collision cascades. The resulting material defects cause degradation of the thermal and mechanical properties of the constituent material. Nanocrystalline materials, characterised by their small grain size, have been proposed for reactor applications due to their high grain boundary density. Grain boundaries are thought to act as sinks for irradiation defects, increasing the lifespan of reactor materials. FeCr binary alloys can be studied as model materials to understand the behaviour of reduced activation ferritic–martensitic steels likely to be used in the first wall and breeding blanket of future fusion reactors. In this poster, we present the results of our irradiation simulations on ARCHER2 and make comparisons with experimental data. LAMMPS is used to run collision cascade simulations, mimicking irradiation damage in nanocrystalline Fe and FeCr. For comparison, initially pristine cells are also irradiated. ARCHER2 was vital for these simulations as it allowed for large cell sizes that capture collective physical effects whilst also enabling large dose simulations to explore longterm evolution. Grain growth is observed for all nanocrystalline cells, and a lower final dislocation density for initially nanocrystalline cells when compared to pristine cells. Chromium does not appear to alter the grain growth rate. Remarkably the growth rate from MD simulations, closely matches that observed in XRD measurements of ion-irradiated, nanocrystalline Fe and FeCr discs. These results reveal key mechanisms underlying nanocrystalline material evolution under irradiation, providing guidance for developing next-generation radiation-resistant steels.
4 Wake Dynamics of Floating Offshore Wind Turbines
Dylan Green, University of Oxford
The platform motions of floating wind turbines can elicit large-scale unsteady responses in the downstream wake behaviour. This has consequences for farm-scale power production and fatigue loading on turbines. As such, optimal design for future floating offshore wind farms requires an improved understanding of the complex mechanisms that drive these unsteady flows. This challenge is further complicated by the turbulent sheared inflow typical of offshore environments.
In this study, we investigate the relationship between floating turbine motion-induced wake dynamics and offshore wind conditions using high-fidelity numerical simulations. A precursor simulation is used to develop a representative sheared turbulent flow. Subsequently, the flow is injected into a successor simulation wherein an actuator line model provides a virtual representation of the turbine. This methodology enables a high-resolution view of the unsteady fluid mechanics, such as discrete blade effects and the subsequent wake dynamics, while maintaining computational feasibility. Addressing this multi-scale problem demands significant computational resources and highlights the role of high-performance computing in advancing technologies critical to net-zero energy goals.
This submission is aimed at ARCHER2 users engaged in large-scale computational fluid dynamics, turbulence modelling, and those interested in and working on renewable energy solutions. The study highlights the use of high-fidelity, computationally intensive simulations to address multi-scale problems, showcasing the capabilities that are relevant across computational science and engineering.
5 Wind farms, understanding the gravity of their atmospheric interactions
Timothy Rafferty, University of Oxford
With the growth of offshore wind farm capacity in the North Sea, the challenges that come from the increased interaction with large scale atmospherics must be understood. Notably the newest turbines, which now reach up to hundreds of metres, can trigger Atmospheric Gravity Waves (AGWs). The buoyancy driven AGWs can cause a redistribution of the pressure field in and around a wind farm altering the flow that the turbines interact with. Whilst it has been shown this will alter farm efficiency and power distribution around the farm, it is not known if the AGWs will impact the flow on the turbine scale.
To evaluate the impact of AGWs on turbine wake dynamics and blade loading, Large Eddy Simulations (LES) of turbine rows were carried out using an actuator line method for turbine representation. Through comparisons to cases without AGWs, it was found the waves could change the turbine wake recovery profile. Simulations also showed turbine spacing within a farm is correlated to the wave strength. To accurately model the North Sea conditions for AGWs required simulations with domains spanning 10s of kms and solvers which could capture turbulent shear layer dynamics, thermal profiles, Coriolis forces and unsteady flow dynamics. Therefore, with the findings in this work also demonstrating the need for a more complete understanding of AGW impacts, it also shows the necessity of the ARCHER2 computational resources for reaching the government’s net zero energy goals.
6 Sampling a North Atlantic Tipping Point: Rare-Event Algorithms in UKESM on ARCHER2
Valeria Mascolo, NCAS, Department of Meteorology, University of Reading
Greenhouse gas emissions are warming the climate. A highly uncertain consequence is the crossing of climate tipping points, where parts of the Earth system undergo rapid, self-amplifying changes that are irreversible on societal timescales. Here we investigate tipping behavior in the North Atlantic subpolar gyre, a circulation feature south of Greenland that redistributes heat, freshwater, and nutrients. A sustained weakening would alter ocean stratification and deep-water formation, with impacts for marine ecosystems and European climate. A central bottleneck in quantifying tipping timescales, impacts and uncertainty is data scarcity: transitions are rare in observations and under-sampled in conventional climate-model ensembles. This limits data-hungry AI/ML approaches, while long high-resolution simulationsneeded to capture processes that trigger transitions are prohibitively expensive. We therefore employ rare-event algorithms, which oversample low-probability transitions by adaptively reweighting and branching ensemble members, generating statistically meaningful samples of rare outcomes at lower cost than direct simulation. We are planning a programme of rare-event sampling experiments in UKESM, designed for execution on ARCHER2. Because the relevant transitions are rare, the approach requires many parallel multi-year coupled simulations with branching and extensive analysis, making it exceptionally demanding computationally. ARCHER2’s national-scale performance and ability to sustain large ensembles are central to the project, enabling us to move beyond undersampled ensembles toward meaningful estimates of SPG transition likelihoods and pathways. We will describe the proposed simulation plan and report initial implementation and performance tests.
7 CAKE: Better connecting and coordinating global Digital Research Infrastructure
Nick Brown, EPCC, University of Edinburgh
CAKE aims to better connect the UK’s Digital Research Infrastructure (DRI). Based on the observation that there is a lot happening in the UK, but it is not always well coordinated, this network aims to act as a conduit to help ensure that teams can more effectively work together, share results, and showcase their expertise both nationally and globally. This poster at the ARCHER2 celebration of science will enable us to share with the ARCHER2 community the existence of CAKE and how we can help them undertake knowledge exchange, highlight some of the wider activities in the UK DRI programme that they could leverage, some of the on-going CAKE initiatives such as the EDI working group, and lastly to share some of our current opportunities for funding including knowledge exchange fellowships and support for undertaking placements.
8 Investigating electron dynamics in the transition from the weak laser-field to the strong laser-field regime
Amy Davis, Queen’s University Belfast
The R-matrix with time-dependence (RMT) code was developed at Queen’s and is a leading method for the solution of problems in attosecond science. Laser-driven electron dynamics evolve on the attosecond timescale, where one attosecond is 10^-18s. We use ARCHER2 to numerically investigate laser-driven electron dynamics. We are interested in how these electron dynamics change when transitioning from the weak laser-field to the strong laser-field regime. To do this we use RMT to investigate how the interaction of a laser-field with helium changes as the laser intensity is increased. We obtain electron spectra which show evidence of new ionization pathways. With RMT we can see how these new pathways are shaped by resonances that are only accessible in strong fields.
9 Development of a turbulence generator based on improved synthetic eddy method (iSEM) and volumetric source term
Jundi He, Department of Mechanical and Aerospace Engineering, The University of Manchester
A handy toolkit that generates user-specified turbulent flow in large eddy simulations (LES) has been developed, which is based on the improved synthetic eddy method (iSEM) and implemented via a volumetric source term. It is capable of generating turbulence within a user-specified box region anywhere in the flow domain. Via a temporarily and spatially varying source term, instantaneous turbulent flow is generated downstream of the box. It is able to generate instantaneous turbulent flow at a user-specified mean velocity and Reynolds stress tensor. Compared with the improved synthetic eddy method which generates turbulence and implemented as Dirichlet boundary condition on a plane (iSEM-P), the turbulent flow generated by the volumetric source term approach (iSEMS) has a much shorter developing distance and allows the use of a larger time step. Moreover, it improves the flexibility and allow user to generate local turbulence at the region of interest, making it particularly suitable for detached eddy simulations (DES). Both methods were tested to generate fully-developed plane-channel flow at a frictional Reynolds number of 395 and the streamwise developments of friction coefficient are compared with periodic LES. The fully-developed turbulent flow generated by iSEM-P experience a laminarisation initially and turbulence is then regenerated over a distance of x/D~5. When the new approach is used, the efficiency is greatly improved and the laminarisation-regeneration process takes a very short, negligible distance.
10 Understanding weather and climate dynamics using high-resolution global cloud resolving models
Evgenij Belikov, EPCC, University of Edinburgh
We introduce the ‘Understanding weather and climate dynamics using high-resolution global cloud resolving models’ project that brings together researchers from CICESE (Mexico), DTU (Denmark) and EPCC (UK) at the intersection of atmospheric science and high-performance computing to establish strategic international collaboration in using high-resolution global cloud resolving models. The project aims to advance understanding of atmospheric physics and dynamics at sub-km scale, whilst developing novel model implementations to analyse wind patterns. In alignment with NERC priorities, selected use cases focus on environmental hazards (e.g. resulting from storms) to improve resilience to extreme weather events and on improving wind resource assessment for wind energy systems which contribute to achieving net-zero targets. We present the results from the physics scheme sensitivity studies at three diverse Mexican locations: La Rumorosa, Merida, and Oaxaca. MPAS-Atmosphere model is used with a publicly available non-uniform mesh configuration with coarsest resolution of 60 km and finest resolution of 3 km over the area of interest. The vertical grid employs 20 m spacing within the first 200 m above ground level, facilitating detailed comparisons of wind speed, temperature, and relative humidity between model output and meteorological station data collected at heights ranging from 20 to 80 m above ground. This work lays the foundation for further collaboration moving to finer mesh resolutions and investigating performance of GPU-enabled MPAS to leverage next generation computing infrastructure, along with MPAS-LES and 3DPBL schemes for analysing the wake effects and atmospheric interactions of wind farm clusters.
11 The structural dynamics and barocaloric response of organic ionic plastic crystals
Rong Wei, University of Edinburgh
Refrigeration systems account for 17% of global electricity consumption. As global temperatures rise, the demand for sustainable cooling technologies grows. Conventional refrigerants are highly efficient but have significant environmental impacts, necessitating greener alternatives. Barocaloric materials, which undergo large entropy and temperature changes under pressure, offer a promising solid-state solution. Organic ionic plastic crystals (OIPCs) exhibit multiple solid-solid phase transitions that can be driven by pressure, making them attractive barocaloric materials. OIPCs have tunable interionic interactions, and their characteristic phase transitions from highly orientationally disordered to ordered phase allow for large entropy changes. However, the microscopic mechanisms governing their phase behavior remain unclear. (Cyanomethyl)trimethylammonium hexafluorophosphate ([N111CN][PF6]) undergoes a four-step disorder-order transition with notable thermal hysteresis. Understanding the nature of the disorder in its high-temperature phase is crucial for optimising its barocaloric performance. This research compares ordered models of the disordered phase to explore its structural flexibility. Ab initio molecular dynamics simulations are employed to explore how the dynamic changes during the phase transition. Phase I is more disordered than other three phases, and the disordered events are likely result from the rotation of the anions [PF]6-. These insights will advance the understanding of OIPCs as next-generation solid-state refrigerants. Furthermore, once a reliable metric has been established to quantify such disorders, large scale parallel simulations will be conducted to generate data to train a machine learning model.
12 Controlling multi-region multi-layered parallelisation for R-matrix with time-dependence, double ionisation
Martin Plummer, STFC Scientific Computing
To ‘view’ electrons in an atom or molecule, we must capture their motion on a time-scale comparable to their interactions. Ultrashort laser pulses can image and control electrons in motion. Free-electron laser facilities operating in VUV, XUV and X-ray regimes have opened up new approaches to investigate ultra-fast dynamics on femtosecond and sub-femtosecond timescales. The complex physics induced by high-energy laser pulses requires close interaction with accurate theory to interpret experimental findings.
R-matrix with time-dependence (RMT) is the world’s most sophisticated code for correlated laser-driven multielectron dynamics in atoms and molecules driven by arbitrarily polarised pulses, including relativistic effects in atomic systems. A major extension treats double-ionisation phenomena: two interacting electrons may be ejected from an atom simultaneously. This massively increases complexity and computational workload. Physical space divides into three regions. The first surrounds the nucleus, utilizing a basis expansion for fully correlated multi-electron effects. The single-ionisation and the double-ionisation regions, with respectively one and two electrons far from the nucleus, utilize finite-difference grids.
Each region adopts numerical schemes requiring distinct multi-layered MPI/OpenMP parallelisation and communication to manage flow of information between regions. Performance issues with the third region restricted efficient operation to near 1500 cores.
A complex additional nested MPI layer in the new region was introduced in an eCSE project to improve performance and match sophistication in the other regions. We present progress on reliable ‘recipes’ for systematic distribution of tasks and threads between regions to achieve (near) optimal load-balancing and performance strategies, crucial for large-scale ARCHER2 calculations.
13 Conformational Isomerism Tunes Refrigeration Potential in Metal Organic Frameworks
Chi Cheng (Cecilia) Hong, University of Edinburgh
As the effects of climate change become more apparent in the form of rising global temperatures, the necessity for temperature control become more significant. Current refrigerants are low cost and work extremely efficiently but these materials possess very high global warming potentials. A greener alternative is a solid-state approach via barocaloric materials which exhibit large adiabatic temperature and isothermal entropy changes upon compression and decompression cycles. Metal-organic frameworks (MOFs) are porous, crystalline materials that are highly modular due to the many metal and linker combinations possible. Notably, certain MOFs exhibit the breathing effect wherein the adsorption and desorption of guest molecules are accompanied by a large reversible volume change. From this, they have been recognised as potential barocaloric materials. Moreover, we have found orientational disorder in the linkers which result in conformational isomers that can tune the barocaloric performance of the frameworks. My research use ab initio molecular dynamics to further understand the effects of the conformational isomerism on the properties of the MIL-53-fum framework series. Grand canonical Monte Carlo simulations allow us to isolate, for the first time, the entropic contributions of guest adsorption from the volumetric or conformational aspects. This reveals the role of breathing MOFs in refrigeration applications and inform the design of functional materials.
14 DRI-focussed training for research facilitators and teams (DRIFT)
Laura Moran, EPCC, University of Edinburgh
Digital Research Infrastructure (DRI) is now ubiquitous in research. It has revolutionised the way that we conduct research, pervading all aspects of the research lifecycle. From the use of large-scale computing for machine-learning and big data, to the management of experiments for record keeping and reproducibility, research is powered by the software and hardware that have become vital for modern researchers. This requires new skills and forms of research collaboration, as research supported by DRI aims at larger scales and interdisciplinary collaboration. This poster covers our work on our three main objectives: supporting research facilitators working within communities that could benefit from DRI offerings so that they can efficiently assist researchers understand its advantages; training research teams in skills that benefit collaboration, team science, and inter- or multi-disciplinary working; and community engagement with a wider range of research institutions and disciplines focussing on those without existing direct access to institutional HPC services. These objectives are being achieved through development of courses and other content for both research facilitators and research teams, community engagement through DRI awareness roadshows (including small demonstration HPC clusters), and the extension and improvement of the open-source Gutenberg training platform. This will improve the way that the expertise required to efficiently and effectively use DRI is developed within the research community.
15 Council Highly coarse-grained polarisable water models for mesoscopic simulations
Benjamin Speake, UKRI Science and Technology Facilities
Modelling micro- and mesoscale thermodynamic and transport properties of soft condensed matter hinges upon its representation. This is especially relevant for polar solvents such as water, since these require effective representation of their dielectric nature as driven by molecular charge distributions and molecular network structuring. The dielectric nature of a medium leads to complex phenomena such as local polarisability response and restructuring near interfaces in reaction to changes in local charge distributions. Inclusion of such phenomena when using larger-than-atomistic techniques such as coarse-grained molecular dynamics (CG-MD) and dissipative particle dynamics (DPD) is still an open question, to which we provide a novel way to consider and justify the necessary and suitable coarse-graining level, enabling us to measure new polar CG models’ performance against that of an underlying atomistic model. We polarise our previous non-polar nDPD water model to prepare it for use in simulations of liquid electrolytes as well as solvated organic membranes and measure its fitness to serve as a dielectric medium by comparing its properties to those of the TIP3P water model, while simultaneously observing changes to properties already represented well by the non-polar model.
16 From Kernels to Solvers: Performance-Portable High-Order PDE Simulation in Nektar++
Diego Renner, Imperial College
Nektar++ is an open-source spectral/hp element framework for the numerical solution of partial differential equations (PDEs), widely used for high-order simulations in computational fluid dynamics (CFD) and related fields. To fully exploit the capabilities of modern heterogeneous high-performance computing (HPC) systems, including graphical processing unit (GPU)-accelerated platforms, the framework is undergoing a major redesign aimed at performance portability and long-term maintainability. The new code enables efficient execution on NVIDIA, AMD, and Intel GPUs, while retaining high-performance CPU implementations single instruction, multiple data (SIMD) vectorisation. This is achieved via a unified backend architecture supporting CUDA, HIP, SYCL, and SIMD operations, enabling device-independent implementation of high-level solver logic while allowing hardware-specific optimisation of performance-critical building blocks such as elemental operators. Building on results previously presented for elemental kernels, this work reports recent progress on optimised global operators, including linear system solvers, evaluated using industry-standard Center for Efficient Exascale Discretizations (CEED) Bake-off Problems (BPs). We additionally demonstrate emerging multi-GPU capabilities and apply the operator-based design to the solver level by refactoring the unsteady advection and compressible Euler solvers around a unified operator interface within the redesigned framework. The resulting solver architecture cleanly separates numerical formulation from execution backends, enabling a single code path to target serial, vectorised, and GPU platforms. These developments represent an important step towards enabling scalable, portable, and production-ready high-order simulations on current and future heterogeneous HPC systems.
17 An investigation of Powersched on ARCHER2
Kevin Stratford, EPCC, University of Edinburgh
Powersched is a tool developed by HPE which aims to reduce the energy usage on compute nodes by dynamically balancing the energy consumed against application performance. We present some initial investigations of using Powersched, and how it changes application performance and energy usage compared with the standard operational configuration on ARCHER2.
18 Hybrid scaling of particle suspension simulations
Kevin Stratford, EPCC, University of Edinburgh
The aim of this ARCHER2 ESCE project is to provide an effective GPU implementation of the particle suspension functionality of Ludwig, a code for complex fluids. In the context of ARCHER2, the use of an thread abstraction means that hybrid MPI/OpenMP performance is also of interest. We describe the approach and present some recent results.
19 Computational modelling of Neon High-Harmonic Generation
Holly Lavery, Queen’s University Belfast
High-Harmonic Generation (HHG) is a process to generate coherent high-frequency XUV light using a visible-light laser. The process starts by removing an electron from its valence shell. Once the electron has escaped the shell, the electric field produced by the pulse causes the electron to accelerate. After acceleration, the electron recombines with the original atom, releasing the kinetic energy it gained during acceleration as a high energy photon. This creates a spectrum of high-order odd harmonics. We investigate HHG from Ne+ ions. These ions have been proposed as a source of HHG light seen in experiment, but these experiments have proven hard to replicate. We make use of the R-Matrix Theory with Time Dependence (RMT) approach to model High-Harmonic Generation of Neon. We make use of separate RMT calculations to model different ionization pathways, and apply our recently developed density matrix approach to accurately model the spin-orbit correlations present between states. These correlations should manifest themselves in a time dependence of the HHG yields.
Full study on HHG requires a detailed description of the atomic structure, making these calculations computationally expensive. We therefore require the use of High-Performance Computing Facilities such as ARCHER2 for this work.
20 Scalable Pipeline for Analysis of Resting-State EEG Spectral Biomarkers in Aging and Neurodegeneration
Amir Miresmaeili, University of Edinburgh
Resting-state EEG offers a powerful tool for studying the aging brain and can help identify neurodegenerative changes. The resting-state signal consists of periodic components (alpha-band power) and aperiodic activity (1/f slope). Decomposing these two elements provides features that can act as markers of cortical health. Prior research has shown that healthy older adults typically have a flatter aperiodic slope (a sign of increased neural noise) along with slowing of individual alpha frequencies. Parkinson’s Disease patients show similar spectral alterations, although the specific patterns may differ between groups. Moving these laboratory findings into clinical practice requires computational resources capable of processing large datasets with complete reproducibility. We applied our analysis pipeline to the ds002778 OpenNeuro dataset, which contains 93 EEG sessions from healthy older adults and Parkinson’s patients recorded across multiple visits. By using ARCHER2, we developed a standardised workflow that extracts physiologically meaningful biomarkers through Welch Power Spectral Density estimation and SpecParam (FOOOF) parameterisation. Deploying on ARCHER2 established a reproducible, scalable framework ready to extend to large-scale population studies, where parallel processing of thousands of subjects becomes essential.
21 Re-engineering FLAMENCO, a Low-Mach Fully Compressible Steady Turbulent Flow Solver, with OPS-DSL for Performance Portability across Diverse HPC Architectures
Ashutosh Shankarrao Londhe, Queen’s University Belfast
FLAMENCO (FLow Analysis and Modeling ENvironment with COmbustion) is a wellestablished high-fidelity solver used for the simulation of turbulent reacting flows with detailed chemistry. The code solves the Navier–Stokes equations augmented with multi-species quasiconservative advection equations using a structured finite-volume formulation. FLAMENCO employs fifth-order accurate spatial discretisation and a second-order extended stability Runge–Kutta time-integration scheme, enabling the resolution of both low- and high-Mach-number flow features, including shocks and turbulence. Parallelism is achieved through MPI-based domain decomposition, and the code has been demonstrated to scale efficiently to multi-billioncell simulations on tens of thousands of CPU cores on leadership-class supercomputers.
In this talk, we present ongoing research exploring the re-engineering of FLAMENCO using the OPS domain-specific language (DSL) with the aim of enabling performance portability across modern multicore CPUs and GPU-accelerated systems. OPS provides a high-level abstraction for structuredmesh applications, allowing the automatic generation of parallel code for multiple backends while significantly reducing the complexity of architecture-specific performance tuning. This approach enables domain scientists to target diverse hardware platforms without requiring detailed expertise in parallel programming models. We discuss the design considerations, challenges, and methodology involved in porting a large, mature Fortran-based combustion code such as FLAMENCO to OPS, including data layout, stencil representation, and preservation of numerical fidelity. This work lays the foundation for future performance and scalability studies of OPS-enabled FLAMENCO on heterogeneous HPC architectures and aims to facilitate efficient large-scale simulations of complex combustion phenomena such as flame–wall interactions, jets, and stratified reacting flows.
22 The Atlantic Meridional Overturning Circulation: Exploring sensitivity to model resolution using ARCHER2
Dan Hodson, NCAS, University of Reading
The Atlantic Meridional Overturning Circulation (AMOC) is an important ocean current that transports warm water poleward, maintaining the temperate climate of the North Atlantic. However, as rising greenhouse gas concentrations warm the climate, the AMOC is projected to weaken. This reduction in heat transport may have profound consequences, including colder winters and disrupted global rainfall patterns. Despite this, the exact mechanism and magnitude of this weakening remain poorly understood. Additionally, standard climate models have a comparatively coarse spatial resolution that fails to capture the observed structure of ocean currents of the northern Atlantic. Modelling studies suggest that the AMOC’s sensitivity is highly dependent on how well these features are resolved. To accurately simulate these processes and bridge the gap between models and reality, high-performance computing (HPC) platforms like ARCHER2 are necessary for the high-resolution, long-term simulations required We present two sets of experiments using both standard and high resolution versions of the HadGEM3-GC3.1 climate model to explore how the AMOC responds to a changing climate: i) CO2 Quadrupling simulations to examine the potential for a rapid AMOC collapse and ii) simulations examining how the atmosphere variability (the North Atlantic Oscillation - NAO) on drives long-term variations in the AMOC.
23 Bayesian Fitting of Quasiparticle Interference
Alexandra Munro, University of St Andrews
Quasiparticle interference (QPI) imaging via scanning tunnelling microscopy (STM) is a powerful technique to probe the low-energy electronic structure of quantum materials. QPI can probe both occupied and unoccupied states and operate at temperatures well below 1K, making it invaluable for studying low-energy electronic excitations in quantum materials and allowing determination of the superconducting order parameter. However, interpretation of QPI data is non-trivial. Extracting physically meaningful parameters, such as hopping amplitudes in tight- binding models or spin-orbit coupling strengths, requires fitting theoretical models to experimental QPI patterns. This fitting is typically performed manually, a process that is both time-consuming and prone to subjectivity. Bayesian inference would provide an assumption free method to fit QPI data. Specifically, using a Markov Chain Monte Carlo (MCMC) framework to estimate model parameters directly from experimental QPI datasets.
24 The Art of the Exercise: Development of Parallel Programming Exercises for EPCC Courses
David Henty, EPCC, University of Edinburgh
Many HPC training courses have associated practical examples, either distributed as working source code for students to compile and run or given as exercises for students to practice writing parallel programs. Designing a good exercise can be challenging as it needs to: be straightforward enough to complete in a reasonable time but complicated enough to illustrate real HPC techniques; be simple enough to explain but be relevant to genuine applications in computational science; produce output that is interesting and engaging but which clearly shows any runtime bugs; scalable to a range of problem sizes to illustrate parallel performance characteristics. Over the years we have used examples from a wide range of areas including cellular automata, image processing, pollution transport via convection-diffusion, predator-prey simulation, percolation of oil through rock, fluid dynamics and sandpile modelling. Despite the wide range of application areas these have all shared the same core characteristic of using a 2D regular grid with nearest-neigbour interactions, e.g. “up, down, left, right minus four times the middle”. In this poster we will explain why this is such a good core computational pattern to use for training exercises, and give the various pros and cons of each particular example. I will also explain why I think the recently developed sandpile simulation is my best example so far, and show the winner from a recent student competition to develop the best sandpile visualisation.
25 Cross-platform GPU Implementation of OpenFOAM Using Only ISO C++ standard parallelism
Yiyun Raynold Tan, STFC Hartree
In this work, we present our port of OpenFOAM to GPUs using the C++ standard parallel execution model (stdpar) introduced in ISO C++17. With a minimally intrusive strategy—primarily replacing serial loops with stdpar constructs—we offload the full computational workload of typical CFD simulations to multicore and manycore architectures. This approach is vendor agnostic, maintains a single unified codebase, and can be integrated easily into the main OpenFOAM release. We demonstrate performance using the icoFoam and simpleFoam solvers across four test cases: the 3D lid driven cavity, 3D conical diffuser, HPC motorbike, and drivAer automotive geometry. Experiments were conducted on a range of NVIDIA and AMD systems, including CPU only, hybrid CPU–GPU, and unified memory CPU–GPU configurations. Measured speedups relative to a fully populated 32 core CPU socket range from 0.4× to 7.7×, depending on boundary condition complexity, turbulence modelling, and solver type. Details of the porting methodology and performance results are provided.
26 Workflow Automation for Reliable Data Transfer and Storage Between ARCHER2 and JASMIN
Juan Carlos Juan Carlos, NCAS
Large-scale simulations executed on ARCHER2 produce substantial data volumes that must be transferred, archived and managed reliably across national research facilities. Manual data movement introduces operational risk, delays downstream analysis, and increases the likelihood of storage bottlenecks. This work presents a production workflow that automates the end-to-end data movement pipeline from ARCHER2 to the JASMIN data facility and onward to long-term Elastic Tape storage. Implemented using Cylc8, the workflow detects newly generated datasets via external triggers, coordinates secure transfers using Globus, verifies successful archival through the Near-Line Data Store (NLDS), and performs controlled clean-up of intermediate storage.
Operational resilience is a primary design objective. The workflow incorporates configurable polling intervals, retry mechanisms to tolerate transient network failures, and structured logging to support traceability and debugging. A modular configuration allows the workflow to be reused across modelling activities and adapted to differing data management requirements. Automating cross-facility data handling reduces time-to-archive, alleviates storage pressure on ARCHER2, and improves the reliability of scientific data preservation. This work highlights the role of workflow automation in supporting sustainable data management practices for Tier-1 supercomputing environments and provides a practical approach for handling the increasing data demands of contemporary simulation workloads.