Invited Speakers

Abstract: Prominent trends in the exascale era and beyond that will affect scalable I/O include: (1) massive concurrency heterogeneity of processing units, (2) large volume of data produced by scientific simulations, observations, and experiments, and (3) a memory and storage hierarchy with heterogeneous devices. These trends impact the performance of parallel I/O, which is a critical mechanism for storing and retrieving data on supercomputing systems. Toward handling this impact, various optimization strategies and novel storage technologies are in development. This talk will review the state of the art and our ongoing R&D in data management techniques targeting upcoming supercomputing systems. The topics include: I/O software stack for storing and retrieving data to and from parallel file systems, I/O optimizations applied in several ECP applications, and ongoing research in I/O libraries such as HDF5 and next-generation user-level object storage, called Proactive Data Containers (PDC).

Bio: Suren Byna is a Staff Scientist in the Scientific Data Management (SDM) Group at Lawrence Berkeley National Lab. His research interests are in scalable scientific data management. More specifically, he works on optimizing parallel I/O and on developing systems for managing scientific data. He is the PI of the ECP funded ExaIO project, and ASCR funded object-centric data management systems (Proactive Data Containers - PDC) and experimental and observational data management (EOD-HDF5) projects.

Abstract: Modern system architectures include multiple tiers of storage organized in a hierarchy. The goal is to mask the I/O gap between compute nodes and remote storage. However, this adds complexity to the end user resulting in under-utilization of these specialized I/O resources. In this talk, we will demonstrate how the existence of multiple tiers of storage, often organized in a hierarchy, can help mitigate the I/O bottleneck. We will present recent developments in two NSF-funded projects designed for and around multi-tiered storage.

Specifically, we will present Hermes, a heterogeneous aware, multi-tiered, dynamic, and distributed I/O buffering system. Hermes aims to remove the complexities associated with multi-tiered storage environments. It offers a simple, yet powerful, buffering API that abstracts the existence of tiers of storage. Further, the Hermes adapter layer is designed to support existing legacy I/O APIs (POSIX, STDIO, MPIIO) transparently to the user via interception. Hermes approaches its first public beta release and we are proud to demonstrate the developments in this area. Further, we will discuss ChronoLog, a new distributed and tiered shared log store. This project is motivated by the activity (or log) data explosion we experience today. This data production stems from the proliferation of modern sensors, IoT devices, web activity, mobile and edge computing, telescopes, as well as by computer-generated traffic (e.g., process synchronization, system utilization monitoring, error debugging, etc). ChronoLog uses physical time to provide total ordering on a log as well as auto-tiering across multiple storage tiers to seamlessly scale the capacity of a log. We will highlight how a distributed shared log can be used to build several other services that require a source of strong consistency, fast appends and “commit” semantics, and transactional isolation.

Bio: Anthony Kougkas is a Research Assistant Professor and the Director of I/O Research in the Scalable Computing Software Laboratory at Illinois Institute of Technology. He earned his PhD from the Computer Science Department of Illinois Institute of Technology, advised by Dr. Xian-He Sun, and his dissertation is titled Accelerating I/O using Data Labels: A Contention-Aware, Multi-Tiered, Scalable, and Distributed I/O Platform. He holds a B.Sc. in Military Science and a M.Sc in Computer Science, both received in Athens, Greece. His research is focused in Parallel and Distributed systems, Parallel I/O optimizations, HPC systems, and Key-Value Store solutions. He currently manages several NSF-funded projects related to distributed and parallel I/O that leverages multi-tiered storage. He works closely with Argonne, Lawrence Livermore, Lawrence Berkeley, and Sandia National Laboratories. Previous to his PhD, Anthony was a military officer serving in several positions in the Army and has extensive experience as an intelligence officer, operations and security, public relations, and personnel development and training.

Abstract: There is commensurate growth in expectations about what can be achieved with this wealth of data and computational power. To meet these expectations with available expertise requires new data and software engineering and management frameworks, new data architectures, and big data processing techniques that make it easier to reliably formalise data-driven methods to extract and analyse information and translate them into actionable insights. Furthermore, it also requires new advanced methods to improve the adoption, sustainability, searchability and reusability of those data-driven methods and other scientific methods/software from and to the scientific communities. This talk will focus on the work that I am doing on developing new advanced information processing technologies and interfaces to extract knowledge from data and software to accelerate: 1) scientific discovery; 2) scientific software adoption, reproducibility, automation, parallelisation, orchestration, and software component re-usability. This includes the development of new scientific data processing workflows/frameworks and programming abstractions, software feature extraction, and scalable and adaptive optimisation algorithms among others.

Bio: Rosa Filgueira, PhD, is an Assistant Professor at the School of Mathematical & Computer Science of Heriot-Watt University since March 2021. Rosa research work focuses on investigating new advanced data-driven frameworks, libraries and architectures, which aim to reduce the effort required to develop data-driven applications at scale that can be run in heterogeneous environments, while ensuring the reproducibility of results and hiding the underlying hardware, software, data complexities. Prior to her position at Heriot-Watt University, she worked as a Research Fellow at EPCC (University of Edinburgh), Research Assistant at the School of Informatics of the University of Edinburgh, and as a Senior Data Scientist at the British Geological Survey (BGS). Dr. Filgueira received BSc and MSC from the University of Deusto (Spain), and PhD from University Carlos III of Madrid (Spain).