SCCS Colloquium - Nov 28, 2019

From Sccswiki
Jump to navigation Jump to search
Date: November 28, 2019
Room: 00.08.053
Time: 15:00 - 16:00

Sebastian Walter: Deep Learning for Parameter Reduction on Real-World Topography Data

Master's thesis introduction talk. Sebastian is advised by Anne Reinarz and Lukas Krenz.

Working with large amounts of topographical data can be challenging in many ways, where trying to reduce the parameter space is probably amongst the biggest of such. For further processing of the data, e.g. to run physical earthquake or tsunami simulation models, one would ideally sample from a much lower dimensional parameter space and, thus, greatly reduce computational effort. The scope of the Thesis is to explore several approaches to achieve this reduction, i.e. Autoencoders, Variational Autoencoders, Generative Adversarial Networks, and compare their performance.

Keywords: Parameter Reduction, Autoencoder, Variational Autoencoder, VAE-GAN

Bruno Miguel: A Distributed Actor Library for HPC Applications

Master's thesis submission talk. Bruno is advised by Alexander Pöppl.

The diversity of high-performance computer systems and application requirements harden the generalizability of software development. For these systems, where performance is critical, abstractions costs are unbearable. But, on the other hand, handcrafting code for every single requirement becomes complex, time-consuming and costly. In this thesis, we present an actor programming model library for high-performance applications running in distributed and shared-memory systems. The library aims to decouple computation and communication logic within finite-state machines entities and to release the pitfalls of parallel designs. We explore well-known standards such as MPI and OpenMP to provide broad cross-compatibility and close to native performance. Its usability and performance are explored on complex structured and unstructured mathematical models, and its results compared with other well-known programming models for HPC, such as the BSP and PGAS. By proxying a tsunami-generated simulation model, we achieve notable scalability on both weak and strong scaling performance tests. Our library allows programmers to focus on algorithms and data-dependencies specifications rather than low-level parallelization constructs and system load-balancing.

Keywords: HPC, NPC, MPI, OpenMP, BSP, SWE