Parallel Computing Training Session 2023

Date: 6th March 2023
Location: CSB Training Room 02.017

Parallel computing is a key technology supporting high-performance computing (HPC) and data analytics. The goal of this module is to provide an overview of parallel computing and introduce attendees to prevailing programming models. The expected outcome of this module is that participants will have an understanding of the different languages and approaches used in this domain and that they will be able to construct simple parallel programs using the discussed languages.

Theoretical lectures and practical training will be dispensed by Hans Vandierendonck, Romain Garnier, Syed Ibtisam Tauhidi, Amir Sabbagh Molahosseini, and Zohreh Moradinia. A general overview of High performance computing will be presented in the morning by Romain Garnier and Amir Sabbagh Molahosseini following by lab classes using practical examples with OpenMP. In the afternoon Syed Ibtisam Tauhidi and Zohreh Moradinia will present how to efficiently handle data analytics with Hadoop and Spark.

RAPID: ReAl-time Process ModellIng and Diagnostics: Powering Digital Factories

This projects aims to develop algorithms for real-time processing for analytics in digital factories. A particular use case is the design of hardware circuits, where silicon can easily be damaged during manufacturing. The production defects that arise negatively affect yield and/or quality of the devices. Uncovering these defects, and how they may be mitigated through tunable process parameters, is a demanding process, especially considering the high production rate and voluminous metrics that are collected.

The project considers the design of sketching algorithms, transprecise computing and their efficient implementation on modern high-throughput hardware such as graphics processing units.

RAPID is sponsored by EPSRC.

Project members: