<font size=5 color=blueyellow>**Note for ISC24 HP Meeting**</font>

[TOC]

## Agenda for ISC24
- ==new ideas from these training topics==

---

## ==May 13==

### 09:00-10:00, Conference Keynote: Beyond Exascale Computing
- https://app.swapcard.com/widget/event/isc-high-performance-2024/planning/UGxhbm5pbmdfMTgyMzc4OQ==
- Katherine Yelick delivered a keynote talk titled **Beyond Exascale Computing**, addressing the challenges and opportunities of post-exascale computing for scientific community.
- Yelick also presented the findings of the US National Academies report that reviewed the hardware and software technologies that are likely to be relevant in this new era of HPC.
- more reading materials:
- https://www.isc-hpc.com/press-releases/conference-keynote-ponder-post-exascale-era.html
- https://www.hpcwire.com/2024/05/15/isc-2024-keynote-high-precision-computing-will-be-a-foundation-for-ai-models/

### 11:00-12:00, Highlights of the 63th TOP500 List
- The 63rd installment of TOP500 list was revealed in coordination with the kickoff of ISC 2024 in Hamburg, Germany.
- The Frontier system at ORNL retains its spot as the most powerful machine on the list.
- In addition, the Aurora machine at the Argonne Leadership Computing Facility in Illinois, USA, continues to occupy the 2nd position on the list.
- The rest of the Top 10 remained largely unchanged other than ==the addition of the new `Alps` machine from the Swiss National Supercomputing Centre in Switzerland. This system debuted at the No. 6 spot on the list and achieved an HPL score of 270 petaflops.==
- more materails for TOP500
- https://www.hpcwire.com/2024/05/13/top-500-aurora-breaks-into-exascale-but-cant-get-to-the-frontier-of-hpc/
- https://top500.org/lists/top500/2024/06/highs/#:~:text=This%20is%20the%2063rd%20edition,way%20into%20the%20Top%2010

### 13:00-14:00, High Performance Software Foundation BOF
- https://isc.app.swapcard.com/event/isc-high-performance-2024/planning/UGxhbm5pbmdfMTgyNjc4NA==
- As the use of HPC becomes ubiquitous in scientific computing and digital engineering, and as AI use cases multiply, more and more data centers deploy GPUs and other accelerators. Open source software projects have been critical for enabling this innovation, but their development and user base must be sustained.
- The High Performance Software Foundation (HPSF) tends provide a neutral home for HPC software projects and to help steward their ongoing development.
- HPSF will sustain the HPC software ecosystem through technical projects, events, training, and promotion, and it will facilitate working groups to help HPC projects collaborate on common goals that benefit the ecosystem as a whole.
- This BoF kicked off the official launch of HPSF at ISC 2024, aiming to give ISC attendees a chance to interact with HPSF's founders, as well as initial member organizations and software projects.
- The general introduction of HPSF, its mission, and its governance structure, as well as some future plans was presented.
- more reading materials
- https://www.linuxfoundation.org/press/linux-foundation-announces-the-launch-of-the-high-performance-software-foundation
- https://www.hpcwire.com/2024/05/14/linux-foundation-announces-the-launch-of-the-high-performance-software-foundation/
- check the note for another BoF session: OpenHPC Community BoF

### 14:15-15:15, HPC Outreach: Telling Our Stories of HPC in Science
- https://isc.app.swapcard.com/event/isc-high-performance-2024/planning/UGxhbm5pbmdfMTgyNjgwNA==
- Outreach is essential to sustainable growth of any community, especially fields like HPC, which few people are exposed to in their everyday lives.
- HPC Outreach is essential to enthusing young minds about computational science, informing the public and growing the community, yet many institutions don’t have sufficient funding or staff effort to support outreach activities.
- Storytelling is a key technique for communicating effectively, especially between people with diverse perspectives. Using storytelling to explain how research is done and why it matters can be a powerful way to build interest. Well-crafted HPC stories are vital to diversify and grow the community.
- The goal of this BoF is to collect and discuss best practices around showcasing successful HPC science stories. The main content of this BoF includes:
- how can we maximise the impact and increase the reach of our message
- How can we communicate complex ideas in a simple to understand way
- How can we raise the public awareness of HPC and attract new talent
- Seven topics were designed for group discussion
- 1, 2, and 6 are quite interesting and are essentially relevant for us to organize workshops
- discussion materials are available at https://drive.google.com/drive/folders/1V38liYW8M-dOdclpUBrZnVAJR5NFXSvc
- Karina talked about her experience in organizing workshops on quantum computing, which is very relevant to the quantum computing workshop delivered by ENCCS

### Poster session

### 16:35-17:35, OpenHPC Community BoF
- check the note for the HPSF BoF in the morning
- https://isc.app.swapcard.com/event/isc-high-performance-2024/planning/UGxhbm5pbmdfMTgyNjgwMQ==
- OpenHPC is the largest open source ecosystem for building supercomputers, and it provides a community-driven stack of common ingredients to deploy and manage Linux-based HPC clusters.
- OpenHPC is in use by hundreds of organizations throughout the world including academic institutes, non-profit organizations, government labs, and commercial entities.
- In this BoF, speakers from OpenHPC Technical Steering Committee provided technical updates from the project and near-term roadmaps.
- Relevant topics on OpenHPC conventions and packaging, requesting additional components and configurations, and general future trends were discussed at this BoF.
- ==OpenHPC *vs.* HPSF==

### Exhibition session about EuroHPC and EuroCC
- Stepas talked about the (training) activities in NCC Lithuania
- Simon from ICHEC (Irish Centre for High-End Computing) presented the training arrangements from NCC Ireland and multiple training projects he was involved.
- ==ENCCS can also prepare some slides and deliver an introduction at the exhibition session==

### Others

---

## ==May 14==

### 09:00-10:30, Programming for an Increasingly Heterogeneous HPC World
- multiple presentations about applications of HPC programming in natural science
- Kokkos - Performance Portability for Everyone
- Christian Trott, SNL
- Achieving Performance Portability has become critical to keep HPC applications and libraries maintainable in the face of growing hardware diversity
- Kokkos has been on the forefront of enabling it, and has demonstrated its ability to efficiently leverage the premiere exascale architectures
- This talk provided a short status update for Kokkos support on the latest HPC architectures, discussing future development directions, and also provided an update on the Kokkos Team's plans to grow the user community as well as the developer base of the Kokkos Ecosystem by transitioning Kokkos to the Linux Foundation.
- ==Software for Quantum Computing: How to (Not) Re-invent the Wheel==
- Robert Wille, TU Munich
- Quantum computers are one of the most promising new technologies which are currently investigated.
- With physical realizations already available to a broader audience and several potential applications on the horizon, this raises the question how to efficiently design corresponding quantum computing solutions.
- As for classical systems, software is key for that. But can we simply re-use established software from the classical realm? Or do we have to start from scratch for quantum computing?
- This presentation provided releant answers to these questions addressing that we do not have to re-invent the wheel---but that a 1:1 re-use of classical software also won’t do the trick.
- Discussions are exemplified by current developments towards Munich Quantum Software Stack and with software tools from Munich Quantum Toolkit (https://www.cda.cit.tum.de/research/quantum/).
- 
- Memory Symphony: Orchestrating Heterogeneity for High-Performance Computing
- Ivy Peng, KTH
- 
- Memory subsystem, as a pivotal element within HPC systems, influences both cost and performance.
- Despite persistent challenges such as the memory wall and memory bandwidth wall, ongoing advancements in new memory technologies, including non-volatile memories and high-bandwidth memories, present heterogeneous memory systems as a promising solution for realizing large-scale memory subsystems while meeting requirements of cost, capacity, and performance.
- Heterogeneous memory often consists of multiple memory tiers exhibiting different characteristics.
- In recent years, heterogeneous memory systems have evolved from HBM-DRAM systems to later byte-addressable NVM-DRAM systems, and with the latest advancements in CXL-interconnected disaggregated memory tiers, significantly expanding the landscape.
- Yet, while hardware-managed caches have simplified programmability, their efficacy wanes when confronted with complex memory access patterns and large memory footprints, as seen in real-world applications.
- Thus, a growing focus shifts towards software-based data placement strategies, where a synthesis of workload insights and system-level understanding becomes crucial for performance optimization.
- This talk delivered by Prof. Peng from KTH presented the memory utilization and patterns on existing leadership supercomputers, brings insights into the future design of heterogeneous memory systems, and then delves into approaches that harness the synergy between workload characteristics and system-level attributes for addressing the challenge in programming heterogeneous memory.

### 10:00-11:00, Arm in HPC: Where We Are Today and Where We Are Heading Tomorrow
- Arm technology has increasingly become a mainstream choice for HPC due to its promise of higher efficiency, density, scalability, and a broad ecosystem of software.
- Arm expansion in the data center started in 2018 with Arm Neoverse, a set of infrastructure CPU IPs designed for high-end computing. The Arm-based Fugaku supercomputer, #1 in the June 2020 Top 500 ranking for 2 years, retained a leadership position over multiple years.
- 
- This event has been a wake-up call for the HPC community.
- The ecosystem is finally mature enough to begin diversifying beyond x86 and re-create a vibrant multi-architecture HPC ecosystem as it was more than a decade ago.
- This BoF aimed to bring together a growing community of leaders and practitioners who decided, either on-prem or in the cloud, to adopt Arm-technology for their current or future production systems.
- A new HPC cluster, `Alps`, hosted by CSCS (Swiss National Supercomputing Centre)
- 

### Lots of light talks and presentations at the exhibition area

### Research Paper Session: Programming Models for Heterogeneous Systems
- Porting HPC Applications to AMD Instinct MI300A Using Unified Memory and OpenMP
- AMD Instinct MI300A is the world’s first data center accelerated processing unit (APU) with memory shared between AMD “Zen 4” EPYC cores and third generation CDNA compute units.
- A single memory space offers several advantages: i) it eliminates the need for data replication and costly data transfers, ii) it substantially simplifies application development and allows an incremental acceleration of applications, iii) is easy to maintain, and iv) its potential can be well realized via the abstractions in the OpenMP 5.2 standard, where the host and the device data environments can be unified in a more performant way.
- This article provided a blueprint of the APU programming model leveraging unified memory and highlight key distinctions compared to the conventional approach with discrete GPUs.
- paper link: https://ieeexplore.ieee.org/document/10528925/authors#authors
- Programming Model Extensions for General-Purpose Processing-In-Memory
- The performance of many applications is limited by available memory bandwidth. One approach to improve the performance of such memory-bound applications is to move the computation closer to the required data.
- Processing In Memory (PIM) integrates computational units directly with the memory. To enable PIM technology in widely used programming models, the authors proposed extensions to OpenMP and OpenACC, two examples of directive-based programming models, as well as SYCL.
- The extensions are designed to be portable across many existing and future parallel computing devices and platforms, making PIM technology widely available.
- By using AMD MI100 GPU with PIM-enabled HBM2 memory, the authors observed a performance improvement of 1.2-2.1 times for representative examples and a real high-performance computing (HPC) application compared to the same GPU without PIM-enabled memory.
- paper link: https://ieeexplore.ieee.org/document/10528942
- Workload Scheduling on Heterogeneous Devices
- Hardware accelerators have become the backbone of many cloud and HPC workloads, but workloads tend to statically choose accelerators leaving devices unused while others are oversubscribed.
- The author proposed a holistic framework that allows a computational kernel to span across multiple devices on a node, as well as multiple applications being scheduled on the same node.
- This work sharing and co-scheduling framework allows kernels to be migrated between devices, expand to span more devices, or contract to fewer devices.
- Experiments on a CPU+GPU+FPGA platform indicated speedups of 2.26X over different applications and up to 1.25X for co-scheduled workloads over baselines.
- Besides performance, a major contribution of our work lies in ease of programmability with a single code base compiled and runtime controlled across three vastly different execution devices.
- paper link: https://ieeexplore.ieee.org/document/10528933

### 14:30-15:30, ==Quantum Computing - Status of Technologies==
- https://app.swapcard.com/widget/event/isc-high-performance-2024/planning/UGxhbm5pbmdfMTgyMzg1Ng==
- The trapped-ion quantum processors at Quantinuum
- Quantinuum has constructed two generations of QCCD (quantum charge-coupled device) quantum processors.
- These processors use trapped-ions for qubits and sympathetic cooling, and shuttling operations to achieve high-fidelity gating operations on individual qubits and between any pair of qubits – making them fully-connected.
- This talk went over efforts to rigorously benchmark the performance of these machines, highlighting their strengths and weaknesses.
- A brief survey of efforts toward near-term quantum advantage and quantum error correction, and some some technology development efforts aimed at scaling these processors and the implications for future devices were also delivered in the presentation.
- Future of HPC: Integrating quantum with massively parallel computing
- As quantum computing systems continue to scale in size and quality, and error resilience approaches start to enable interesting computational regimes in what we call the era of quantum utility, the integration of quantum with massively parallel computing becomes critical to unlock the full potential of both technologies in a way that exceeds the capabilities of either one alone.
- This integration is poised to provide a rich environment for experts to experiment and optimize resources in quantum algorithms and applications.
- Given the limitations in efficiently emulating quantum applications to find the most optimal implementations, direct interaction with evolving quantum hardware becomes essential.
- This talk presented the state of the art of quantum computing and will touch on some architectural ideas towards the integration of quantum and traditional HPC systems.
- Superconducting Quantum Computers in the NISQ era and beyond
- Quantum computing teams worldwide are dedicated to showcasing utility—an essential milestone before achieving quantum advantage.
- This involves demonstrating the reliability of NISQ quantum computers in executing computations at a meaningful scale.
- Simultaneously, quantum computing providers are tasked with crafting compelling roadmaps toward fault tolerance.
- This presentation presented how the team at IQM spearheading innovative solutions to tackle these pivotal aspects of quantum computing advancement.

---

## ==May 15==

### 9:00-10:00, ==HPC-QC Initiatives in and for Europe – Where is Hybrid Quantum-HPC Headed?==
- https://isc.app.swapcard.com/event/isc-high-performance-2024/planning/UGxhbm5pbmdfMTgyNjc5Mg==
- 
- In recent years, Quantum Computing (QC) systems has gradually matured and started to migrate from laboratories into data centers with the intent to merge into the HPC ecosystem and to provide acceleration capabilities for particular workloads.
- Currently, several parallel efforts drive this process within the European ecosystem.
- This BoF brought together representatives of the five most important initiatives defining and supporting Europe’s leading efforts for HPC-QC integration to discuss their approaches and perspectives:
- The EuroHPC JU has recently awarded six European hosting sites, with more planned, to seat quantum computers with the express intent to develop and provide hybrid HPC-QC instruments.
- 
- Horizon Europe, an additional European Commission instrument, has developed a large portfolio of research projects, including the Quantum Flagship.
- 
- The Quantum Flagship is responsible for preparing a recent version of the Strategic Research and Industry Agenda (SRIA) document.
- 
- The European Quantum Industry Consortium (QuIC) coordinates European Quantum companies and European industry striving to boost their global competitiveness through the use of quantum technologies.
- 
- Finally, the European Quantum Systems and Software Summit, EQS3, a series of community building and aligning events, strives to unite quantum and HPC practitioners at a deep technical level around hybrid quantum-HPC integration and at many levels in the data center environment.
- 
- This BoF brings together representatives from all five initiatives for a conversation on their respective motivations, status, and plans, followed by a discussion with the HPC community, especially the targeted European user community in the audience.

### 10:05-11:05, ==Developing a Sustainable Future for HPC and RSE Skills: Training Pathways and Structures==
- https://isc.app.swapcard.com/event/isc-high-performance-2024/planning/UGxhbm5pbmdfMTgyNjc4OA==
- Despite the quantity of existing training materials, acquisition and development of the specialist skills required for HPC is not straightforward enough to address the needs of the growing and diversifying HPC community.
- To address this, the HPC teaching and training ecosystem needs to mirror the growth and diversification of the HPC community and related technologies.
- This BoF will explore the arising new requirements in order to identify further entry points to HPC and RSE skills development.
- It will also support the building of well-defined learning pathways that more accurately represent the aims of the user/learner community and changing technology landscape.
- The learning pathways provide two major benefits, first they encourage a shift from teaching topics in a linear manner toward concept based learning pathways that better align with adult learning models.
- 
- 
- 
- 
- Secondly, they are the next step in developing guidelines to make the training content more FAIR - findable, accessible, interoperable and reusable.
- 
- Developing interoperable, citable and persistent training materials is key to creating personalised learning pathways that directly correspond to the training needs and job requirements of HPC community members, especially in the exascale era.
- This session gathers experts from several HPC and RSE projects and initiatives from across the world to lead the community discussion on the development of novel learning pathways.
- It will build on previously collected data and an understanding of community perspectives to advance the development of approaches to learning in the HPC community. We encourage everyone involved or interested in HPC training to attend.
- check more details at these links:
- https://www.universe-hpc.ac.uk/
- training pathways: there is PDF document (figure below) and I will try to find it out later
- 
- ==trainig materials==: https://www.universe-hpc.ac.uk/resources/training-material/
- The pathway exercise is super informative, we can design training events and workshop topics following similar network structure

### ==11:10-12:10, Current Topics in EuroHPC Training==
- https://app.swapcard.com/widget/event/isc-high-performance-2024/planning/UGxhbm5pbmdfMTgyNjc4Mw==
- 
- Representing a significant share of the Education & Training community in Europe, this BoF aims at providing a user forum to discuss new emerging trends in contemporary HPC training across Europe.
- Focused on, but not limited to, specific goals of the EuroCC2/CASTIEL2 projects and their efforts in HPC training, a selected list of key subjects is summarized as follows:
- Training Baseline Orthogonal to similar efforts currently ongoing in Europe, the training baseline seeks to specify a set of HPC training courses particularly suited to customize industrial user profiles. Characteristics are modular components, community-derived content, variability in all its building blocks, associated certification scheme (e.g. micro-credentials) and in general a co-design effort of European scale
- Assuring the Quality of Online Training Programs The quality assurance of training programs needs to be considered from the first step, to ensure that the needs of the intended audience are correctly identified, their learning needs are met and that they are fully engaged from the start. We will look at ways that this can be achieved to maximise user engagement to ensure that the intended learning is understood and retained
- (<span style="background-color: lime">Me and Sven will talk about the successful story for quantum computing</span>) Quantum Computing in Training: current situation & success stories The training program on quantum computing has been initialized in academia and industry seeking to balance the frontiers of research and individual studies. Based on past training events, we seek to have collaborative efforts between academia, industry, and research institutions, in order to foster knowledge exchange and trigger new initiatives.
- Continuing last year’s initiative our main interest lies again in collecting various viewpoints from all the people actively involved in HPC training (both providers as well as user communities) including training providers from European Competence Centres (NCCs) and Centres of Excellence, industrial training centres and those from various supercomputing sites.

### Lots of light talks and presentations at the exhibition area

---
---