Digital Twin Development for Taiwan's tokamak (FIRST) research: Multi-Scale Plasma Simulation, AI Equilibrium Reconstruction, and 3D Visualization

This talk presents recent progress by our team in developing a digital twin for Formosa Integrated Research Spherical Tokamak (FIRST), Taiwan’s premier nuclear fusion tokamak project. Currently in Phase 1 of development, our research integrates simulation, state reconstruction, and high-fidelity visualisation. We have established a temporal baseline for tokamak plasma evolution and implemented PLaNet [1] to accelerate magnetic equilibrium reconstruction. To capture multi-scale phenomena, we investigate critical plasma instabilities using targeted numerical codes: JOREK [2] for magnetohydrodynamic (MHD) dynamics such as vertical displacement events (VDEs), and GTC [3] for micro-instabilities like ion temperature gradient (ITG) modes. Furthermore, to bridge theoretical simulation with immersive visual analytics, we have rendered 3D JOREK simulation outputs within an NVIDIA Omniverse environment. These milestones establish the computational infrastructure needed for Phase 2 development as we prepare for FIRST’s initial plasma discharge.

[1] Bonotto, Matteo, Domenico Abate, and Leonardo Pigatto. Fusion Engineering and Design 200 (2024): 114193.

[2] M. Hoelzl et al, 2021 Nucl. Fusion 61 065001

[3] Z. Lin, T. S. Hahm, W. W. Lee, W. M. Tang, and R. B. White, Science 281 (5384), 1835-1837 (1998).

About the Speaker

Dr Yi-Cheng Chuang is a researcher at NCHC specialising in magnetic control fusion. He holds a Ph.D. from William & Mary, where he studied neutral particle transport in tokamaks. Today, his work centres on advancing Phase 2 of digital twin development. When he’s not modelling plasma environments, he enjoys hiking local trails and playing board games.

Addressing Fusion Energy Challenges through Multiscale Simulation and AI

The development of practical fusion energy presents significant scientific and engineering challenges, from understanding complex plasma behaviour to managing interactions between high-temperature plasmas and material surfaces. Addressing these challenges requires modelling physical processes across vastly different spatial and temporal scales, while also extracting useful insights from data generated by fusion experiments.

This presentation highlights research at A*STAR’s Future Energy Acceleration and Translation (FEAT) Centre, where high-performance computing and artificial intelligence are being applied to advance fusion science and technology. Drawing on research in computational plasma physics and plasma–material interactions, we will illustrate how numerical simulations are being used to understand plasma behaviour and predict material response under fusion conditions. In parallel, machine learning is being applied to extract meaningful information from diagnostic data collected by fusion experiments worldwide. Together, these efforts demonstrate how HPC-enabled simulation and AI can help researchers tackle key scientific and engineering challenges on the path towards practical fusion energy.

About the Speaker

Dr Eddie Chua is Co-lead of Plasma Physics & Diagnostics and a Senior Scientist at A*STAR’s Future Energy Acceleration and Translation (FEAT) Centre. He obtained his Ph.D. in Astrophysics from Harvard University, where his research centred on large-scale cosmological simulations of galaxy formation. His current research focuses on computational plasma physics, including edge and sheath dynamics, plasma–material interactions, and impurity transport. He also leads research in artificial intelligence for fusion diagnostics, integrating physics-based modelling and machine learning to advance predictive capabilities for future fusion energy devices.

From Physics to AI and back

In this talk, we explore how we can use pre-existing knowledge (e.g., physics) to improve AI systems, and how we can possibly extract some knowledge from AI systems. On the first topic, we present a novel physics-enhanced deep learning hybrid method, namely CondensNet, for resolving cloud physics in climate models. On the second topic, we present some results on the use of explainable AI for climate applications. We conclude with an overall perspective bridging the two topics.

About the Speaker

Dr Gianmarco Mengaldo is an Assistant Professor in the Department of Mechanical Engineering and the Department of Mathematics - by courtesy - at National University of Singapore. He is also a member of the Joint Advisory Group for the World Meteorological Organization (WMO), a United Nations (UN) agency. He received his BSc and MSc in Aerospace Engineering from Politecnico di Milano (Italy), and his PhD in Aeronautical Engineering from Imperial College London (United Kingdom). After his PhD he undertook various roles both in industry and academia, including at the European Centre for Medium-Range Weather Forecasts (ECMWF), and at the California Institute of Technology (Caltech). Dr Mengaldo adopts an interdisciplinary approach at the intersection of mathematical engineering, computational physics and AI to study complex systems that arise in various branches of applied science. His current research interests involve (i) integrating domain knowledge (e.g., physics) and AI; (ii) explainable AI, both theoretical and applied, among others. Dr Mengaldo’s main application areas include weather and climate, robotics, and finance.

Token Cartography: principled applications of machine learning techniques to retain scientific interpretability

Unsupervised machine learning, such as clustering and latent reduction, is commonly applied to large-scale scientific data. These methods serve as the bedrock for more complex scientific foundation models. Lacking introspective approaches to access the trained latent spaces, scientific insights that might have been learned cannot be revealed. Furthermore, outlier and extrapolation detection is often not baked into the model, leaving users with outputs that further require confidence assessment. Together, these shortcomings render many existing foundation models unexplainable, significantly limiting scientific elucidation. To address this drawback, we propose a strategy to construct interpretable latent spaces by inducing semantic directions and mapping the space with known chosen scientific descriptors, alongside data regularisation via vector quantisation and feature engineering. This talk will elucidate the strategy, termed Token Cartography, and show an early use case on a massive complex petabyte-scale diffraction imaging dataset, achieved with HPC resources. We believe that this strategy can serve as a conceptual guideline for procedures to maximise extractable information from experimental data but also lay the groundwork for developing explainable scientific foundation models.

About the Speaker

Ervin Chia is a Doctoral candidate of Physics at the National University of Singapore and is part of an international collaboration aimed at understanding anomalous low-temperature properties of the most vital molecule for life on Earth: water. He studies petabyte-scale data collected from fourth-generation light sources, using machine learning methods enabled by HPC resources, to gain insights on the microstructures that form in the brief moments as water crystallises into ice.

Navigating Chemical Space: Techniques for mapping chemistry for molecular identification and reaction path discovery.

Most modern materials discovery AI techniques typically target the prediction of a final product with desirable properties. There is a growing understanding that more information about the chemical processes that result in materials is required in order to ensure these materials are synthesisable, among other desirable information, such as prediction of observables so we can ensure the correct materials are being produced, and behaving as expected.

This presentation will present two bodies of work.

1. Mass Spectroscopy is a standard technique for the identification of molecules in a laboratory setting, particularly for molecular species previously known, and catalogued into libraries. Identification of unknown molecules requires more extensive work, either through use of further molecular analysis techniques (such as Nuclear Magnetic Resonance or Infrared spectroscopy, among others), or computationally predictive techniques that can generate a synthetic library data entry to match a molecule to before the molecule is ever synthesised. This latter approach is particularly attractive when molecules are expected to be dangerous. Our lab had developed a molecular fragmentation and characterisation workflow for the prediction of Mass Spectrum using Density Functional Theory methods, which will be discussed.

2. Reaction Route Maps are a technique of searching chemical space through potential energy surface (PES) minima, and transition states between these minima, these can be labelled as zeroth, and first order saddle points respectively. We have developed the Molecular high-Index Saddle Dynamics (MHiSD) method which seeks to map chemical space by searching up the PES landscape the 2nd, 3rd and higher order saddle-points, and then back down in a branching tree approach. This provides alternative routes across the PES, and additional information on the structure of the PES, ideal for a broader, and more detailed search of chemical space. This technique is general for any molecular modelling method that can generate gradients and hessian information, and early results will be discussed.

About the Speaker

Stephen Dale is an Assistant Professor in the Department of Materials Science and Engineering, and the Institute of Functional Intelligent Materials at the National University of Singapore. His research focuses on the development of density-functional theory (DFT) and its applications in the discovery and characterisation of new materials. Stephen obtained his PhD from the University of California, Merced in 2017 and BSc from the University of Western Australia in 2012. Before joining NUS, he was a Postdoctoral Research Fellow at Dalhousie University, the Australian National University, the University of Sydney, and Griffith University.

Revealing Earth’s Interior with TomoATT: HPC-Enabled Seismic Imaging and Discovery

Understanding subsurface structures is essential for exploring energy resources, assessing geohazards, and supporting underground development. These structures must be inferred from indirect seismic observations by solving computationally demanding inverse problems. This talk introduces TomoATT, an open-source software package for adjoint-state traveltime tomography, and explains how algorithmic advances and high-performance computing enable detailed three-dimensional seismic imaging. I will highlight our ongoing studies across the Japanese archipelago, California, and Singapore, illustrating how TomoATT supports investigations of Earth’s structure at regional and local scales.

About the Speaker

Tong Ping is an Associate Professor at Nanyang Technological University, Singapore. His research focuses on applied & computational mathematics, computational geophysics, and seismic imaging, with applications to energy exploration, geohazards, and subsurface characterisation. His team develops advanced mathematical methods and open-source software, including TomoATT and SurfATT, for imaging Earth’s interior using seismic observations. By combining adjoint-state methods, high-performance computing, and geophysical data, his work connects computational innovation with practical challenges in geothermal exploration and geotechnical engineering.

Combustion simulation at scale: How GPU supercomputing is unlocking large-scale simulations of hydrogen and ammonia flames and detonations

Hydrogen and ammonia contain no carbon and produce no carbon dioxide when burned, making them among the most promising fuels for reaching net-zero emissions. Yet ammonia is difficult to ignite, and an accidental hydrogen release can develop into a highly destructive detonation. Understanding these combustion processes underpins the design of cleaner engines and the safety assessment of fuel storage and transport. Numerical simulation is central to this research, but it is extremely demanding, and large-scale, high-fidelity simulations have long been out of reach on conventional CPU-based supercomputers. This talk shows how GPU supercomputing is changing that. On NSCC's ASPIRE 2A and ASPIRE 2B systems, our team benchmarked and deployed GPU-accelerated combustion solvers and found that a single GPU can do the work of roughly 80 to 200 CPU cores. This has made it possible to simulate ammonia ignition and hydrogen detonation at a scale that was previously impractical. New insights from these simulations will be provided, including practical lessons from running large-scale computations on GPU supercomputers and what this work means for clean energy and safety.

About the Speaker

Dr Wang Yicun is a Postdoctoral Research Fellow in the Department of Mechanical Engineering at the National University of Singapore, working in the group of Associate Professor Zhang Huangwei. He received his PhD in Power Engineering and Engineering Thermophysics from Zhejiang University in 2024. His research covers high-fidelity combustion simulation and combustion modelling, together with artificial intelligence for combustion, including reduced-order models and physics-informed neural networks. His recent work focuses on GPU-accelerated simulation of low-carbon fuels such as hydrogen and ammonia. He has published more than 15 peer-reviewed journal papers in combustion and energy, and serves as a peer reviewer for several journals.

Thinking beyond speed

High-performance computing has always been about speed: faster machines, faster codes, shorter time-to-solution. In this century, energy efficiency and the decarbonisation of the power grid have moved from secondary concerns to central, and arguably existential, constraints. An energy-efficient machine is essential, but it is far from enough. This talk argues that efficiency has to be understood across the whole hardware–software–energy ecosystem, and that the effectiveness of an HPC centre depends on something usually left out of the equation: its users. Not all workloads are equally urgent, and users hold knowledge about deadlines, flexibility and value that no scheduler can infer. I will argue for bringing the user into the loop, and user needs into the cost function that governs how an HPC centre operates.

About the Speaker

Rui Oliveira is an Associate Professor of Fault-Tolerant Distributed Systems at the University of Minho, a senior researcher, and Director of International Affairs at INESC TEC. He serves on the boards of the Portuguese National Center for Advanced Computing (CNCA), the Irish Centre for High-End Computing (ICHEC), and the EuroHPC Joint Undertaking. His research addresses scalable, dependable, and privacy-preserving data management systems, as well as energy-efficient computing workloads. Oliveira has authored more than 100 peer-reviewed publications and has served as coordinator or senior researcher in over 30 nationally and European-funded research and innovation projects. As General Director of the Minho Advanced Computing Center—the precursor to CNCA—he oversaw the deployment and operation of Portugal’s first national high-performance computing system (Bob) and directed the design and funding proposal for the EuroHPC supercomputer, Deucalion.

Hybrid MCMC for Bayesian Inverse Problem, Taking Advantage of Both World - AI Surrogate and Numerical

Many scientific computing and engineering simulation problems can be classified as inverse problem, where the ultimate goal for simulation is to optimise engineering design, parameter calibrations, or field reconstruction from partial observations. However, solving inverse problem can be computationally expensive especially for Partial Differential Equation (PDE) governed problems. Recent advances in Deep Learning (DL)-based surrogate models have shown promising potential to accelerate the solution of such problems. However, trained by the numerical data, AI surrogates may not match the accuracy or preserve the physics as good as the numerical models. In this talk we will present a hybrid method utilising both AI surrogate and high-fidelity numerical models in the Markov chain Monte Carlo method, where we take advantage of the inference speed of AI surrogate models for speedy sampling and high-fidelity numerical models for error correction.

About the Speaker

Juntao Yang is an AI4Science specialist at NVIDIA, part of the NVIDIA AI Technology Center (Singapore). Prior to joining NVIDIA, he has been working in the industry as a simulation engineer. He joined NVIDIA in 2017 through the NVIDIA-NTU Industry Postgraduate program. Now he works with researchers and scientists in the Asia Pacific (SEA+ANZ) region to solve impactful research problems with state-of-the-art NVIDIA technology. His main interest is the convergence of scientific computing and AI, particularly in solving PDE-constrained problems with AI algorithms.

ROCm 10: A deep dive

As ROCm celebrates the 10 year anniversary with the release of ROCm10, a peek into the history and the latest and greatest features offered by ROCm for production grade AI deployments.

About the Speaker

Paul is a solution lead at AMD. He has over a decade of experience in HPC/AI infrastructure including system design and implementation. He plays a crucial role in building AMD's AI/HPC ecosystem across ASEAN region.

From Data Lake to Discovery: Building AI-for-Science Pipelines on ASPIRE 2B’s Data Platform

ASPIRE 2B pairs more than 1,500 NVIDIA H200 GPUs with a data platform that is more than a scratch filesystem. Alongside a single all-flash namespace served over NFS with RDMA, S3, SMB and GPUDirect Storage, the same platform exposes queryable file metadata, transactional and vector database semantics, and event-driven functions — the building blocks of a research data lake. For researchers, this changes what a “storage system” can do: data can be governed, searched, embedded and acted on where it lands, without copying it between silos.

This talk is a practical, demo-driven walk through two end-to-end AI-for-Science pipelines built on that platform, framed by a healthcare data framework that sets out the development process — data preparation and governance, model development, clinical validation and monitoring — and the personas each stage needs, from health data architects to clinical domain experts. Both pipelines are shown as recorded walkthroughs, paused and narrated at each step. The first is a medical-imaging workflow: images are ingested into the datastore, tags drive governance and trigger pipelines, image metadata is queried together with structured records, a MedGemma model runs inference and writes its findings back to the database, and every access is captured in the audit log. The second is an open-source genomic RAG engine: FASTQ files trigger NVIDIA Parabricks alignment and variant calling, variants are annotated against ClinVar and AlphaMissense, embedded into a vector store, and queried through hybrid search with LLM synthesis and citations, with NVIDIA BioNeMo MolMIM and DiffDock closing the loop into candidate-molecule generation. Both show the same pattern that leading facilities such as NERSC are adopting for instrument-to-insight workflows: ingest → govern → query → infer on one platform.

The talk is deliberately user-focused rather than product-focused. It closes with what these capabilities mean for Singapore’s research community: how a research group can bring its own pipeline — in healthcare, genomics or any other data-intensive domain — onto the ASPIRE 2B data lake, the governance questions to settle first, and where scale-out inference and large-model checkpointing are heading next.

About the Speaker

Dr Jen Chang “James” Chen is Field Technical Director – HPC at VAST Data, where he supports national supercomputing centres and AI infrastructure programmes across APJ. His work spans storage architecture and benchmarking for large-scale procurements, application I/O optimisation on production systems, and the data-platform requirements of emerging AI-for-Science workloads.

James brings nearly two decades of experience across academia and industry. Before joining VAST Data in 2024, he was a Senior HPC Solutions Architect at Amazon Web Services (2021–2024), helping ASEAN and APJ customers run HPC workloads in the cloud, and held HPC and data-centre roles at AMD and GIGABYTE. From 2017 to 2020 he was a Senior Research Fellow at the National Supercomputing Centre (NSCC) Singapore, working on system architecture, performance benchmarking and cloud bursting for the national HPC platform. Earlier, at Academia Sinica’s Institute of Atomic and Molecular Sciences in Taiwan, he designed and operated an HPC facility over several years.

He holds a PhD in Computational Materials Science from National Taiwan University, and his published work includes best practices for molecular dynamics on AWS Graviton3E and scaling financial simulations on AWS.

HPC for Quantum Utility: From Digital Twins to Autonomous AI Scientists

Quantum computing is progressing rapidly, yet realising practical quantum utility remains challenging due to limited quantum resources and the complexity of developing reliable quantum applications. In this talk, I will discuss how high-performance computing (HPC) can accelerate the path toward quantum utility through two complementary directions. First, HPC enables us to build digital twins of quantum systems, using large-scale simulation and AI to learn predictive models that reduce the need for costly quantum experiments. Second, HPC provides the computational foundation for Omni, our autonomous AI framework for quantum error correction, enabling AI agents to design, simulate, verify, and improve quantum workflows at scale. These results indicate a new avenue about how HPC, AI, and quantum computing can work synergistically to make useful quantum computing more accessible and accelerate scientific discovery.

About the Speaker

Yuxuan Du is an Assistant Professor at Nanyang Technological University (NTU), jointly appointed to the College of Computing and Data Science and the School of Physical and Mathematical Sciences. His research focuses on autonomous quantum science, quantum error correction, and fundamental algorithms for quantum computing. His work has been published in top-tier journals and conferences across physics and computer science as first or corresponding author, including Nature Reviews Physics, Nature Communications, Physical Review Letters, Physical Review X Quantum, npj Quantum Information, IEEE Transactions on Information Theory, and IEEE Transactions on Pattern Analysis and Machine Intelligence. He has been listed as a Top 2% Scientist Worldwide in the field of AI and Physics, based on citation indicators published by Stanford University and Elsevier. His research has been featured by major media outlets such as MIT Technology Review and Phys.org. He has served as an Area Chair for NeurIPS and ICLR. He serves as an Editor of Neural Networks and Quantum Machine Intelligence, and has regularly held Area Chair roles for ICML, NeurIPS, and ICLR.

Designing Hybrid Quantum–Classical Systems Before Deployment: Insights from HybridTwin

Hybrid quantum–classical systems bring together HPC resources and quantum processors, but their performance depends on more than the capabilities of either component alone. Communication latency, resource contention, and scheduling decisions can all change how a hybrid workload behaves. How can we evaluate these interactions before committing to a system design?

In this talk, I will introduce HybridTwin, a multi-fidelity digital twin framework for architecture-aware exploration of hybrid quantum–classical systems. It connects early-stage system simulation with execution-level analysis and real-device evaluation. Using HybridTwin, we examine how CPU–QPU communication latency changes the relative performance of quantum, classical, and quantum simulation. We then show how, as the number of concurrent jobs grows, waiting time and scheduling increasingly shape end-to-end performance. The talk concludes with a comparison of job-level and task-level scheduling and discusses what these findings mean for designing scalable hybrid computing systems.

About the Speaker

Dr Tao Luo is Senior Scientist and Head of the High Performance Computing (HPC) Chapter at the Institute of Advanced Intelligence and Computing (IAIC), A*STAR, Singapore. His research focuses on high-performance computing systems, AI infrastructure, scientific computing, computer architecture, and quantum computing, with an emphasis on enabling advanced computing capabilities for scientific discovery and industrial innovation. He leads initiatives spanning HPC platforms, AI systems, and large-scale computing infrastructure, and works closely with academia, industry, and national stakeholders to strengthen Singapore’s computational research ecosystem.

Dr Luo has led and contributed to multiple national and industry-funded programmes in AI and HPC. His work has been published in leading venues including Nature Communications, Communications of the ACM, IEEE TPDS, IEEE TCAD, SC, ICLR, NeurIPS, ECCV, AAAI, etc. He was an Associate Editor of IEEE TNNLS and currently serves as an Associate Editor of IEEE TCAD and has held numerous leadership roles in the international research community, including Track Chair of HPCAsia 2026, Program Chair of NVMSA 2024, and General Chair of NVMSA 2025.

AI-Guided Tensor-Network Simulation for Large-Scale Quantum Computing on HPC Systems

Tensor-network methods offer a promising approach for simulating quantum circuits beyond the practical limits of conventional state-vector simulators. However, scaling tensor-network simulations to large quantum circuits presents significant challenges, including contraction-order optimisation, high intermediate memory requirements, and efficient execution on modern GPU-accelerated HPC systems.

This talk presents our work on combining artificial intelligence with tensor-network algorithms to improve the scalability of large-scale quantum simulations. In particular, we explore reinforcement learning for contraction-order optimisation and memory-aware slicing techniques for efficient multi-GPU execution. Using representative quantum-circuit workloads, we demonstrate how AI-guided optimisation can be integrated with HPC methodologies to enhance computational efficiency and scalability. The presentation will highlight how the convergence of AI, tensor-network methods, and high-performance computing can help address the growing computational demands of quantum computing research.

About the Speaker

Dr Vinitha Balachandran is a Senior Scientist at the Institute of Advanced Intelligence and Computing (IAIC), A*STAR, Singapore. She obtained her PhD in Physics from the National University of Singapore and has extensive experience in computational quantum physics. Her research focuses on quantum many-body systems, quantum transport, and quantum information, with expertise in tensor-network methods and advanced numerical techniques for simulating complex quantum systems. More recently, her work has expanded to the intersection of artificial intelligence and quantum computing, including AI-driven approaches for quantum simulation, optimisation, and large-scale quantum computing.

Co-Designing Data Center Networks and Distributed Systems

Distributed systems traditionally treat the network as a passive substrate, leaving coordination and resource management to end hosts. Programmable networks offer an opportunity to reconsider this division of responsibility. This talk explores how co-designing network mechanisms and distributed systems can overcome bottlenecks in scalability and performance while preserving application correctness.

I will present two systems that illustrate this approach. First, Hydra addresses the scalability limitations of in-network ordering, a technique that simplifies strongly consistent distributed protocols by ordering requests within the network. It replaces a single network sequencer with distributed sequencers, combining loosely synchronised physical clocks with logical ordering to scale throughput while preserving ordering guarantees. Next, Capybara tackles a complementary challenge: dynamically balancing load for stateful services built on TCP. By co-designing a programmable switch, a host network stack, and an application-state migration interface, Capybara migrates live TCP connections in approximately 15 microseconds without disrupting clients, enabling rapid load redistribution across servers.

Together, these systems demonstrate how carefully chosen abstractions between networks and end hosts can make distributed services more scalable and efficient. I will discuss the design principles behind these abstractions and the practical challenges of reconciling network programmability with protocol semantics and deployment requirements.

About the Speaker

Jialin Li is the Sung Kah Kay Assistant Professor in the School of Computing at the National University of Singapore. He finished his PhD at the University of Washington in 2019. His research work has received the ACM SIGOPS Hall of Fame award, and best paper awards at NSDI ’26, NSDI ’15, and OSDI ’14. His current research interests include co-designing distributed systems with modern hardware, data plane operating systems, and distributed AI/ML systems.