Date: May 5, 2022 | 3.00 p.m. (GMT+1)
Speaker: Ricardo Macedo, Research Assistant, Institute for Systems and Computer Engineering, Technology and Science (INESC TEC) and University of Minho
Moderator: Todd Evans, Researcher Associate, Texas Advanced Computing Center (TACC)
HPC infrastructures are long thought as computational powerhouses that enable scientists to conduct massively parallel jobs. However, with the advent of new data-intensive workloads from both scientific and deep learning jobs, the storage performance has become a pressing concern in these infrastructures, due to high levels of performance variability and I/O contention generated by multiple applications executing concurrently. In this webinar, we discuss how to build portable and generally applicable Software-Defined Storage data planes tailored for the requirements of data-centric applications running on modern HPC infrastructures. We demonstrate how to improve I/O performance and manage I/O interference of HPC jobs with none to minor code changes to applications and HPC storage backends.
About the speaker:
Ricardo Macedo is a researcher at HASLab INESC TEC and a 4th year PhD student at University of Minho, advised by Prof. João Paulo and Prof. José Pereira. His research interests include local and distributed storage systems, with particular focus in Software-Defined Storage, programmable and distributed storage, and kernel-/user-space storage technologies. Currently, Ricardo is working on a Software-Defined Storage data plane framework for enforcing end-to-end performance policies in complex storage infrastructures.
About the moderator:
Todd Evans joined TACC in 2013 as a member of the High Performance Computing team. Todd received his Ph.D. in Physics from the University of Illinois at Urbana-Champaign in 2008. He spent the following two years as a postdoc in high-energy physics at the University of Regensburg and then three years as a postdoc in Nuclear Engineering at North Carolina State University. Since joining TACC his research focus has been on application performance monitoring, optimization and analysis.
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.
