Speaker

Jeff Larkin, Director

Session description

Industrial engineering is undergoing a significant transformation, propelled by advancements in accelerated computing, AI Physics, and digital twins. This presentation will detail how these hardware and software innovations are helping to compress engineering design cycles and accelerate research and development.

We will deep-dive into mixed-precision techniques, new ML architectures, emerging quantum computing applications and the evolution of GPU hardware. Through real-world use cases from the automotive, aerospace, and semiconductor industries, this session will provide the HPC community with actionable insights into how AI-driven simulation and accelerated computing is fundamentally reshaping the industrial engineering landscape.

Speaker

Stefano Carrazza - Associate Professor, University of Milan and INFN Milan

Session description

Discover Qibo, an open-source middleware for quantum computing and hybrid-quantum-classical integration. Singapore’s National Quantum Computing Hub (NQCH) is part of the global network contributing to the development of this full-stack option. Qibo supports the entire quantum workflow, from high-level algorithm development and large-scale simulation to low-level hardware control and execution on physical quantum processors. These tools are deployed on the Hub’s superconducting quantum processor. Learn more at nqch.sg

Speaker

Eric Mansfield

Session description

Incorporating Quantum Computing (QC) into High Performance Computing (HPC) environments (commonly referred to as HPC+QC integration) marks a pivotal step in advancing computational capabilities for scientific research. In this talk, we share a firsthand account of integrating a superconducting 20-qubit quantum computer into the HPC infrastructure at LRZ Supercomputing Center in Germany, one of the first practical implementations of its kind. This yielded four key lessons: (1) quantum computers have stricter facility requirements than classical systems, yet their deployment in HPC environments is feasible when preceded by a rigorous site survey to ensure compliance; (2) quantum computers are inherently dynamic systems that require regular recalibration that is automatic and controllable by the HPC scheduler; (3) redundant power and cooling infrastructure is essential; and (4) effective hands-on onboarding should be provided for both quantum experts and new users. By sharing these experiences, we aim to provide a roadmap for other HPC centers considering similar integrations.

Speaker

Akira Sano, General Manager FAE & Business Development

Session description

Supermicro DCBBS builds optimized HPC with Plug&Play services, from custom design to implementation.

Speaker

NAKAMURA HIROSHI

Session description

Redefine the future datacenter via scaling servers and advanced cooling solutions

Speaker

Jason Strawderman, Senior Director, WW Platforms Business Development - Western Digital Corporation

Session description

As AI data pipelines and HPC workloads reshape data-center design, disaggregated storage over high-speed ethernet fabrics is emerging as the foundation of performance, scalability, and efficiency. By aligning low-latency flash tiers for real-time training, inference, and simulation with cost-optimized QLC flash and HDD for model repositories, datasets, and checkpoints, data centers can fluidly balance speed and capacity. This new architecture unifies workflows - accelerating innovation, simplifying scale, and redefining the economics of data-driven discovery. Join us to explore a forward-looking vision for infrastructure innovation and how it’s shaping the future of intelligent workloads.

Speaker

Masahiro Waki,
AI/DX Strategic Business Development Lead
Senior Business Development Manager
NetApp G.K.

Session description

NetApp enables Japan's AI for Science initiatives with secure, integrated, and advanced intelligent data infrastructure.

By leveraging strategic partnerships with leading AI companies like NVIDIA, AWS, Microsoft, and Google Cloud, NetApp provides AI-ready data platforms that simplify complex AI data processes and accelerate scientific research. NetApp's solutions, such as AFX and AI Data Engine, offer performance, flexibility, and control, ensuring efficient, compliant, and secure data handling.

Committed to supporting efforts led by RIKEN, HPCI, and the Japanese government, NetApp facilitates breakthrough discoveries in various fields, including life sciences and material sciences, through innovative data management and hybrid cloud integration.

Speaker

Hiroshi Horii, Head of IBM Quantum Japan

Tatsuhiro Chiba, Senior Technical Staff Member, AI Infrastructure, IBM Research

Session description

We would like to introduce the highlights of IBM's Lunch Speaker Session, Keynote Speakers Session, Invited Session, and Fugaku tour. We will also introduce IBM's initiatives in the fields of IBM Quantum, IBM Vela/AIU and IBM Storage.

Speaker

Toshiaki Hishinuma, Principal HPC Engineer / Director of TechDev Dept., SB Intuitions Corp.

Session description

We introduce Sarashina, Japan’s cutting-edge LLM developed by SB Intuitions. This session explores the infrastructure demands for LLM training and inference, highlighting our innovative solutions to optimize performance and scalability. We discuss our AI-optimized environments across data centers, enabling efficient model training while addressing resource challenges. Furthermore, we demonstrate how these platforms facilitate Sarashina’s lightning-fast inference and seamless integration with third-party services, unlocking new possibilities for enterprise AI applications. Join us to discover how we’re redefining AI infrastructure for next-gen intelligence.

Speaker

Werner Scholz, CTO and Head of R&D Xenon Systems

Session description

Cluster management tools and user interfaces are often the neglected layer between the users, their applications and the cluster hardware. But can a thoughtfully architected HPC cluster software stack enable modern cluster design, regardless of where the hardware lives, and include support for a broad range of platforms and applications from classic scientific codes to modern prebuilt container images?

Learn how XENON Cluster Stack frees one from vendor and/or cloud lock-in and allows for infrastructure flexibility, modular expansion, and easy customisation for heterogeneous clusters. This results in an agile infrastructure that is composable and enables experimentation.

Speaker

Brian Christiansen, VP of Engineering

Session description

Slurm is an open-source workload manager designed to satisfy the demanding needs of HPC, HTC, and AI workflows. SchedMD is the company behind Slurm, working as the sole developer and provider for support, training, and consultation. Together, SchedMD services and Slurm work to enhance site security, reduce risk, increase performance, and boost scalability. Now, Slurm can be combined in a Kubernetes environment with Slinky to provide Kubernetes workloads better access to resources, higher scalability, and more efficient AI worklows.

Speaker

Jani Heikkinen, VP of Business Development

Session description

QUDORA Technologies is a developing full-stack trapped-ion quantum computers based on its proprietary Near Field Quantum Control (NFQC) technology. Current laser-based quantum systems face scaling bottlenecks, high error rates, and significant operational overhead, limiting their path to fault tolerance. QUDORA’s microwave-based NFQC architecture eliminates these bottlenecks, enabling longer coherence times, and higher gate fidelities. At same time QUDORA is building QCaaS capabilities and already provides access to its emulator on cloud.qudora.com

Speaker

Shimon Ben-David, Chief Technology Officer

Session description

Yesterday's data infrastructure can't keep up with the demands and scale of AI, HPC, and other compute-intensive workloads. Join WEKA's CTO Shimon Ben-David to discover how purpose-built AI storage eliminates latency bottlenecks, reduces costs, and unlocks AI innovation at scale.

Speaker

Alvin Cheang, HPC/AI Business Director

Session description

This workshop will provide attendees with insights into solutions designed to support AI workloads and infrastructure strategies for AI-driven, compute-sovereign environments. The focus will be on edge deployments and modular on-premises compute units, aligned with HPC/AI business strategies.

Speaker

Shintaro MOMOSE Ph.D., Senior Professional

Ikuta MASASHI, Senior Professional

Session description

The Next Vector Project by NEC tackles the pressing challenges in high-performance computing (HPC), including energy efficiency, scalability, and accessibility. Building on the proven SX-Aurora TSUBASA vector architecture, the initiative integrates the open-standard RISC-V instruction set to drive innovation and global collaboration. In partnership with Openchip & Software Technologies in Spain, NEC is co-developing a next-generation processor system. The project’s mission is to democratize powerful vector computing, extending its reach from HPC specialists to domain scientists. The presentation also highlights NEC’s cutting-edge AI solutions, such as materials informatics.

Speaker

(Tentative) Thierry Pellegrino, Global Head of Advanced Computing

Session description

(Tentative) AWS gives researchers the power to turn their boldest questions into world-changing discoveries. A couple of success stories with societal benefit realized by AWS advanced computing are introduced in this session, including our future vision towards Quantum Computing.

Speaker

Kimmo Koski, CSC - IT Center for Science

Damien Lecarpentier, CSC - IT Center for Science

Session description

CSC - IT Center for Science is shining a spotlight on the LUMI Supercomputer capabilities and the LUMI AI Factory offering to showcase the opportunities and vision behind combining world-class supercomputing, high-quality data, and top-tier technical and AI expertise to empower Finnish and European innovation.

Speaker

Stephen Wu, Director Partnerships, Digital Research Alliance of Canada

David Tweddell, Interim Vice President Strategy and Planning, Digital Research Alliance of Canada

Session description

As Canada’s national digital research infrastructure organization, the Digital Research Alliance of Canada (the Alliance) has developed a bold vision for a generational investment in AI compute to support over 26,000 researchers across the country. The Government of Canada has committed CA$705M as a new investment into public digital research infrastructure. This investment signals the need for national sovereignty and the consideration for implications of global shifts in geopolitics and supply chain. Learn more about the Alliance’s five pillar AI sovereign compute strategy, which includes a consolidated and scaled up AI compute facility, national services, and national data platform.

Speaker

Takashi Nakagawa TOKYO ELECTRON DEVICE LTD.

Session description

Wafer Scale Engine - one of the world's largest chips, built from a full 12 inch wafer - powers the Cerebras CS-3 to deliver tightly coupled compute, huge on-chip memory, and extreme bandwidth. This 15-minute session explains why generative AI needs purpose-built compute, ensuring data control and overcoming the scaling limits of multi GPU systems. We then show how CS-3 accelerates LLM and HPC workloads and highlight performance and efficiency achieved with wafer-scale integration. Finally, we outline practical paths to evaluate and deploy with TED expertise and support.

Speaker

Michael Granitzer, University of Passau, Germany

Session description

We will present OpenWebSearch.eu aimed to build an open index of the Web on HPC computing resources. Funded by the EU’s Horizon Europe programme, the initiative brings together 14 European research and computing centres to develop a sovereign, legally compliant Open Web Index. This shared index is conceived as a digital commons, openly available worldwide and ready to power search, generative AI, web analytics and domain-specific applications. It will be operated on federated HPC infrastructure available to innovators, researchers and entrepreneurs for research and development.

Speaker

Keisuke Kishida, Offering Management Lead, Quantinuum

Session description

In this session, we will briefly introduce Quantinnum, one of the world's leading quantum computing companies, and its latest model, Helios.

Quantinuum Helios, Powered by Honeywell, is the world’s most accurate general-purpose commercial quantum computer, designed to accelerate quantum computing adoption by enterprises.

With the highest fidelity of any commercial system and a first-of-its-kind real-time control engine, Helios enables developers to program a quantum computer in the same way they program heterogeneous classical computers.

Speaker

Aki Asahara, Fixstars Corporation

Session description

As AI workloads grow increasingly complex and resource-intensive, performance engineering has become a critical discipline for achieving both efficiency and scalability. Fixstars specializes in advanced software optimization across heterogeneous computing platforms. In this session, we introduce Fixstars AI Booster ⸺ a high-performance AI acceleration framework that optimizes model training and inference on GPU clusters through profiling, scheduling, and parameter tuning. We will present case studies demonstrating up to several-fold improvements in throughput and cost efficiency. Attendees will gain insights into practical performance engineering methodologies that drive sustainable, high-impact AI infrastructure for research and industry.

Speaker

Jan Moren

Session description

University labs have traditionally acquired and managed their own research support and equipment. We're today moving toward a centralized model with university-funded *core facilities* providing support and services; and a *common resources* approach to providing shared equipment through the core facilities.

At the Okinawa Institute for Science and Technology, we have implemented the Core Facilities and Common Resource model, and it is the default way to handle research support and purchase and management of big-ticket equipment. We will discuss how this is implemented at OIST, and the benefits that we realize from it.

Speaker

Todd Gamblin - Lawrence Livermore National Laboratory

Damien Lebrun-Grandie - Oak Ridge National Laboratory

Session description

A retrospective of the projects and progress of the High Performance Software Foundation, and a preview of what's coming in the next year.

Speaker

Frank Herold CEO

Philipp Falk VP of Engineering

Session description

Modern HPC and AI workloads are pushing traditional storage systems to their limits. In this session, ThinkParQ showcases how BeeGFS, the leading parallel file system for performance-driven environments, empowers organizations to eliminate data bottlenecks, scale seamlessly, and fully utilize GPU and CPU resources. Learn how BeeGFS accelerates demanding workloads, from deep learning to seismic simulation, through distributed metadata, flexible scaling, and hardware independence. We’ll share customer use cases, best practices, and performance insights that demonstrate how BeeGFS transforms storage into a strategic enabler for next-generation HPC and AI success.

Speaker

April Chen Senior Manager, Lenovo ISG Product Group

Executive Director, Enterprise AI & HPC, ISG ESMB Segment & AI

Session description

The SC777V4 presentation showcased Lenovo’s next-generation Neptune supercomputing tray, delivering extreme performance and energy efficiency in a compact footprint. Featuring the latest technology in the market and 100% direct water cooling, it redefines HPC scalability for AI and scientific workloads, supporting ExaScale ambitions with sustainable, space-saving design.

Speaker

Hiroki Sakamoto

Session description

In the construction of AI HPC data centers, the biggest challenges are achieving advanced integration as a data center while simultaneously reducing costs and shortening construction time. Our company addresses these challenges with a completely integrated concept based on our internationally patented aisle-linked expandable container technology, which almost eliminates the need for construction and installation work.

Speaker

Fumikazu KONISHI, PhD.Eng, Research Center, SAKURA internet Inc.

Session description

SAKURAONE is a production, vendor-neutral GPU supercomputer that tests whether an all-Ethernet fabric running SONiC can meet state-of-the-art AI/HPC demands. Built with 100 nodes, dual Xeon 8580+ CPUs, eight NVIDIA H100 SXM5 GPUs per node, and an 800-GbE leaf-spine RoCEv2 network, it placed 49th on the TOP500. The system delivers 33.95 PFLOP/s HPL and 339.86 PFLOP/s HPL-MxP, with HPCG at 396.3 TFLOP/s. A 2-PB all-flash Lustre file system and a reproducible software stack (Rocky Linux 9, Slurm, CUDA, Apptainer) support workloads. IO500 scored 181.91; MLPerf Training on GPT-3 and Llama-2 shows competitive scaling, validating Ethernet+SONiC at national scale for deployment.

Speaker

Makoto Akagi, CEO

Session description

Are your random numbers truly random?
Discover optimal quantum randomness⸺Ultra-Fast, Efficient, and suitable for HPC.

Speaker

JINNAI Erika, Public Relations Division

Session description

HPCI, Japan’s innovative High-Performance Computing Infrastructure, connects major supercomputers and large-scale storage systems across universities and research institutions via high-speed networks. It includes the national flagship supercomputer Fugaku and other leading systems, providing a flexible computational environment to meet diverse academic and industrial needs.

Through open calls, HPCI welcomes users from around the world, offering various project categories and comprehensive support services to help maximize computational performance.

This session will introduce the HPCI framework, its computing resources, and available user support services.

Speaker

Yoshifumi Takafuji

Session description

As Intel enters a transformative phase, this session outlines a new direction in next-generation computing. One highlight is the newly launched Intelﾂｮ Xeonﾂｮ 6+ Processor. The session covers recent developments in HPC and AI, with an emphasis on sustainable computing. It also explores the design philosophy and architectural strategies behind next-generation platforms, and how Intel is working to improve computational efficiency and scalability in high-performance computing.

Speaker

Gen Shimada, Linaro

Marcin Krzysztofik, Linaro

Session description

Linaro is a global leader in the Arm software ecosystem, driving innovation and performance across the HPC landscape. In this talk, we present Linaro Forge, a cutting-edge suite of debugging and profiling tools designed to simplify application porting between architectures and accelerate performance optimization. Learn how Linaro Forge empowers developers to achieve maximum efficiency on Arm and beyond, supporting the next generation of high-performance computing.

Speaker

Koichi Hyodo

Session description

Ethernet keeps evolving and is expanding its use cases in HPC and AI networking. This session introduces the latest Ethernet technology to scale-up, scale-out and scale-across HPC and AI networking.

Speaker

Cameron Brunner, Head of HPC Development

Session description

We will share our vision for applying Agentic AI to amplify the power of HPC by enhancing its ability to align computational utilization with organizational goals, drive smarter resource management, and maximize the strategic impact of HPC in your organization.

Speaker

Elisabeth Ortega-Carrasco, Solutions architect

Session description

Qilimanjaro Quantum Tech is a full-stack quantum computing company based in Barcelona, focused on analog quantum computation through the design of fluxonium-based quantum chips. These systems are integrated with commercial digital transmon platforms and classical resources in under the same software stack, facilitating to the ecosystem its use and adoption. This approach is deployed at Qilimanjaro facilities creating the first multimodal quantum data center of Europe, which can be accessed through SpeQtrum, its Quantum-as-a-Service platform, and at the Barcelona Supercomputing Center (BSC) within the MareNostrum-Ona hybrid infrastructure.

Speaker

Michael James, Founder and Chief Architect, Cerebras Systems

Session description

AI’s success has driven new massively parallel processor architectures with characteristics distinct from traditional HPC. Cerebras Systems pursued a fundamentally different design point focused on strong scaling, limited only by physical constraints. We built the world’s largest chip⸺a wafer-scale processor packed with compute and memory, a scale requiring 3D power delivery. The architecture keeps application data close, matches IO to computational bandwidth, and achieves strong scaling 1000× beyond contemporary exascale platforms for certain applications.

This talk introduces the Cerebras Wafer-Scale Engine architecture and surveys applications running on it⸺from AI training and inference to PDEs and large-scale simulations. We demonstrate what becomes possible with extreme strong scaling parallelism

Speaker

Satoshi Nakajima, Team Director, Processor Development Division, Fujitsu

Session description

AI is increasingly adopted in HPC, not only for data analysis but also to complement simulations. As workloads grow more complex, Agentic-AI requires real-time adaptability, making CPU-centric infrastructures essential. FUJITSU-MONAKA, a 2nm Arm-based processor launching in 2027, builds on A64FX’s legacy to deliver high performance, energy efficiency, and reliability. Supporting liquid- and air-cooled servers, FUJITSU-MONAKA meets diverse deployment needs and complies with Arm SystemReady and Redfish. This session introduces Fujitsu’s technologies for enabling the convergence of HPC and AI.