HPC Webinar for SMEs: Best practices from Central Europe

HPC Webinar for SMEs: Driving growth through collaboration: Insights from real HPC Use Cases

Join us for an insightful webinar designed to showcase the transformative potential of High-Performance Computing (HPC) through real-world success stories and compelling projects realized with the support of the National Competence Center (NCC). This event aims to highlight the practical applications of HPC, demonstrating its impact and inspiring SMEs and industry partners to embark on collaborative Proof of Concept (PoC) and pilot projects.

In addition to use cases implemented at the Slovak NCC, the webinar will also showcase the expertise and experiences of neighboring competence centers in the Czech Republic and Poland. Our speakers will present their exceptional use cases, illustrating the capabilities and achievements of their NCCs. Our speakers will present their exceptional use cases, illustrating the capabilities and achievements of their NCCs.

This webinar is open to all SMEs and industrial companies; we also look forward to welcoming the members of our HPC ambassadors. Don’t miss this opportunity to learn from leading experts and discover how HPC can drive innovation and success in your projects.

September 4th from 14:00 to 15:30 CET
Registration

The event will be held in English.

Venue: online via zoom

After registration, participants will receive a confirmation email. Please check your Inbox/Spam folder. A few days before the webinar, a calendar invitation to an online Zoom meeting will be sent to all registered participants.

SPEAKERS:

Michal Pitoňák: Slovak National Supercomputing Center, Senior HPC application developer

Short Bio: Doc. Mgr. Michal Pitoňák, PhD. has been involved in the field of High-Performance Computing (HPC) for over 20 years. He was part of the operational team of Slovakia's first supercomputer, Aurel. Since 2023, he has been working at the National Supercomputing Center, where he contributes to the implementation of pilot projects within the National Competence Center for HPC project, funded by the Joint Undertaking EuroHPC.

Tomáš Karásek: IT4Innovations National Supercomputing Center, Parallel Algorithms Research Lab, Czech Republic

Short Bio: Tomas Karasek, Ph.D. is currently the Head of the National Competence Centre in the High Performance Computing of the Czech Republic and head of the Parallel Algorithm Research Lab at IT4Innovations National Supercomputing Center. Previously he spent seven years at several positions at Institute of High-Performance Computing in Singapore. In past he participated in several EC funded projects concentrating on SMEs such as SESAME Net, InnoHPC, and the PRACE SHAPE program. Dr Karasek has received funding for more than 30 projects of contractual research from industrial partners in the past 15 years, and is the author or co-author of more than 20 scientific papers.

Szymon Mazurek: AGH University of Krakow, AI developer, NCC Poland

Short Bio: Szymon Mazurek (PhD. student) is an AI developer with expertise in research and development projects. He has worked with vision, natural language, and time series data using various types of networks, including CNNs, transformers, recurrent, graph, and spiking neural networks. He has extensive experience with the usage of HPC environments in deep learning, both for research and model deployment. Szymon is interested in embedded and embodied AI, especially with the use of neuromorphic platforms and spiking neural networks. Actively conducting research in this domain, he is pursuing a PhD at AGH University of Krakow.

PROGRAM:

14:00 – 14:05 Introduction: Brief presentation of NCC Slovakia

Speaker: Lucia Demovičová: Head of NCC for HPC Slovakia

14:05 – 14:25 Use Cases NCC Slovakia

Speaker: Michal Pitoňák

Case 1: Transfer and Optimization of CFD Workflow in HPC Environment
Short Description: This case study focuses on the transfer and optimization of Computational Fluid Dynamics (CFD) workflows within an HPC environment.
Link to Publication: Transfer and Optimization of CFD Workflow in HPC Environment

Case 2: Anomaly Detection in Time Series: Gambling Prevention Using Deep Learning
Short Description: This case involves using deep learning techniques to detect anomalies in time series data for the purpose of preventing gambling.
Link to Publication: Anomaly Detection in Time Series: Gambling Prevention Using Deep Learning

Case 3: Named Entity Recognition for Address Extraction in Speech-to-Text Transcriptions Using Synthetic Data
Short Description: This case study addresses the identification of entities to extract addresses from transcribed conversations using synthetic data.
Link to Publication: Named Entity Recognition for Address Extraction in Speech-to-Text Transcriptions Using Synthetic Data

Case 4: Measurement of Structural Parameters of Capsules Using AI and ML Techniques
Short Description: This case explores the use of artificial intelligence (AI) and machine learning (ML) techniques to measure structural parameters of capsules.
Link to Publication: Measurement of Structural Parameters of Capsules Using AI and ML Techniques

14:25 – 14:30 Q&A

14:30 – 14:50 Use Cases NCC Czech Republic

Speaker: Tomáš Karásek

Case: Examples of the use of artificial intelligence in solving problems from engineering practice

Description: The AI boom in recent years has not been limited to generative AI or large language models but has also affected other areas, including engineering applications. This talk will present two case studies of using AI methods in energy and transportation. In both cases, an approach where numerical modelling and simulation methods were used to generate the data needed to train AI models. The synthetic data thus generated was subsequently used to train and validate the AI models. In both cases, HPC infrastructure was advantageously used to accelerate computational simulations and train AI models.

14:50 – 14:55 Q&A

14:55 – 15:15 Use Cases NCC Poland

Speaker: Szymon Mazurek

Case: SpeakLeash - Community-Driven Initiative for Advancement of Polish National LLM Ecosystem

Description: Recently, Large Language Models (LLMs) have emerged as a breakthrough technology, offering unprecedented automation opportunities across nearly all domains. However, most of these solutions are developed and provided by a limited group of companies, posing challenges for applications that require specific resources, data confidentiality, and tailored tasks. This talk will explore SpeakLeash, a Polish open research community dedicated to developing LLMs tuned for the Polish language. The group's activities in data gathering and cleaning, model development, evaluation, and serving in an open-source approach will be examined. Additionally, the importance of supporting such initiatives will be discussed, highlighting their necessity for developing resources and know-how to use state-of-the-art technologies.

15:15 – 15:20 Q&A

15:20 – 15:30 Discussion

Follow the NCC social media profiles and get the latest news about training events.