Throughout the second semester of 2021 into 2022, the UT Austin Portugal Program and the BigHPC consortium will be organizing a webinar series to spark discussion among academia and industry about the challenges around High-Performance Computing (HPC) and the major opportunities that lie ahead for both sides.
Apart from session one, which will exceptionally bring together three senior researchers from TACC for an engaging conversation about containerized application performance at petascale, all remaining sessions will take the form of a 30-minute overview talk by an academic or a business expert involved in the BigHPC project, followed by a 10-minute moderated discussion with attendees.
Sessions are intended to provide rather practical insights into HPC-related topics ranging from containerization technologies to scientific visualization, to HPC monitoring and data storage.
This webinar series attests to the consortium’s commitment to knowledge sharing and dissemination beyond its partners and direct stakeholders to convey how collaboration between academia and industry can play an important role in advancing HCP for the benefit of science, societies, and the economy at large.
The BigHPC Project is co-financed by the European Regional Development Fund through the Operational Program for Competitiveness and Internationalisation – COMPETE 2020, the Lisbon Portugal Regional Operational Program – Lisboa 2020 and the Portuguese Foundation for Science and Technology – FCT under UT Austin Portugal.
Check the full agenda below.
Webinar #1: Containerized Application Performance at Petascale
- Date: July 22, 2021 | 3.00 p.m. (UTC+1)
- Speakers: Richard Todd Evans, Manager of HPC Performance and Architecture Group at the Texas Advanced Computing Center (TACC) | Amit Ruhela, Research Associate (TACC) | Stephen Lien Harrell, HPC Engineering Scientist (TACC)
- Moderator: Ricardo Macedo, Researcher, INESC TEC
Containerization technologies provide a mechanism to encapsulate applications and many of their dependencies, facilitating software portability and reproducibility on HPC systems. However, in order to access many of the architectural features that enable HPC system performance, compatibility between certain components of the container and host is required, resulting in a trade-off between portability and performance. In this work, we discuss our experiences running three state-of-the-art containerization technologies on five leading petascale systems.
Webinar #2: Is HPC ready for “Big” Data Storage?
- Date: September 16, 2021 | 3.00 p.m. (UTC+1)
- Speakers: João Paulo, Researcher, INESC TEC and University of Minho
- Moderator: Júlio Costa, Data Analyst, Wavecom
While the computational power of HPC infrastructures is rapidly increasing and is expected to be further improved in the forthcoming years, there has been significantly less research and development on storage solutions that can accompany this evolution. In fact, storage performance (i.e., the access and retrieval of data from applications) is now a major setback for efficiently supporting a new generation of data-centric Big Data applications in HPC. This webinar will address the main data storage challenges that are being brought by Big Data applications deployed at HPC infrastructures. Also, it will discuss possible solutions, backed by recent research work, that can potentially alleviate this pressing storage bottleneck problem.
Webinar #3: Monitoring in the BigData Era
- Date: November 18, 2021 | 3.00 p.m. (UTC+1)
- Speakers: Bruno Antunes, Research Manager, Wavecom
- Moderator: João Dias, Minho Advanced Computing Center (MACC)
Monitoring is one of the main challenges of the BigHPC project for improving the management of HPC data centers and Big Data applications. From the basic application of triggering alarms to alert for malfunction hardware, services, applications, etc, monitoring in the HPC area encompasses much more. Nowadays it is expected to also provide visualization of the collected data, which will allow not only to profile the monitored components but also to help system admins to predict possible malfunction or misbehaved applications. It will also allow users to improve the performance of their applications, targeting the specific points that could benefit from a redesign. In the Big Data era, the monitoring component should be able to filter, from the huge quantity of collected data, the one with meaningful importance to the end-user. In this webinar, we will get an overview of the tools we are trying to develop for the BigHPC project.
Webinar #4: Software Defined Scientific Visualization
- Date: January 20, 2022 | 3.00 p.m. (UTC)
- Speakers: João Barbosa, Researcher, Minho Advanced Computing Center (MACC)
- Moderator: Samuel Bernardo, Software Engineer, Laboratório de Instrumentação e Física Experimental de Partículas (LIP)
This talk will cover new methodologies of scientific visualization in high-performance computing systems specially designed for large-scale scientific visualization that provide a greater level of accuracy and detail.
