Abstract

The primary task in my master thesis is the implementation and simulation of chaotic systems, as well as subsequent evaluation with respect to parameters of the integration methods. For the implementation the programming language Julia (https://julialang.org) is used, which is especially suited for parallel and distributed computations.

Researchers

Principle Investigator: Prof. Dr. Andreas Polze || Contact Author: Joachim Kruth || University of Potsdam, Germany

Abstract

We want to evaluate how page replication as another management mechanism is worth the effort. The experiments are based on a patched linux kernel running on multi-socket NUMA systems. The outcome should be to identify possible workloads and data structures that likely benefit from the replication of pages.

Researchers

Principle Investigator: Prof. Dr. Andreas Polze || Contact Author: Felix Eberhardt || Hasso Plattner Institute, Germany

Abstract

We have developed an agent-based game-theoretic model which promises a good explanation of the structure of real world networks. Previous large-scale experiments revealed that our model fails to produce networks with non-constant diameter. This project aims at simulating variations of the model that allow a more flexible diameter.

Researchers

Principle Investigator: Prof. Dr. Tobias Friedrich || Contact Author: Dr. Pascal Lenzner || Hasso Plattner Institute, Germany

Abstract

We are planning to evaluate our work on PINPOINT (to be published at the Runtime and Operating Systems for Supercomputers workshop at Super Computing 2020) for energy-efficient and economic processing on heterogeneous compute infrastructures in the HPI Future SOC Lab in cooperation with the OSM Group (HPI).

Researchers

Principle Investigator: Prof. Dr.-Ing. habil Wolfgang Schröder-Preikschat || Contact Author: Dr.-Ing. Timo Hönig || Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany

Abstract

In this project, we use machine learning techniques to predict the optimal timing for running a VM live migration.  We use machine learning techniques for live migration cost prediction and datacenter network utilization prediction.  This helps IT admins to be alerted with the optimal timing recommendation when a live migration request is issued.

Researchers

Principle Investigator: Prof. Dr. Christoph Meinel || Contact Author: Mohamed Elsaid || Hasso Plattner Institute, Germany

Abstract

Partial database replication is a query-driven approach to minimize the overall memory consumption of a replication cluster while still enabling a balanced load distribution among nodes. In this Future SOC Lab project, we want to deploy partial data allocations for large database clusters and evaluate their end-to-end performance.

Researchers

Principle Investigator: Prof. Dr. h.c. Hasso Plattner || Contact Author: Stefan Halfpap || Hasso Plattner Institute, Germany

Abstract

Microservices and DevOps are gaining considerable attraction. We would use the requested HPI Future SOC Lab resources to investigate the experimental evaluation of our activities on "DevOps-oriented Load and Resilience Testing for Microservices" and "Automated Cross-Component Issue Classification for Microservices".

Researchers

Principle Investigator: Dr.-Ing. Andre van Hoorn || Contact Author: Dr.-Ing. Andre van Hoorn || University of Stuttgart, Germany

Abstract

Our project contributes to the field of behaviour-based authentication. In the last years we collected a great amount of walking sequences of several people. As this dataset is too large to be processed on a normal machine, we hope to evaluate and improve our authentication model by the support of additional resources.

Researchers

Principle Investigator: Prof. Dr. Christoph Meinel || Contact Author: Vera Weidmann || neXenio GmbH, Germany

Abstract

In this project, we study mechanisms for integrating hardware accelerators in virtual machines and cloud infrastructures. Exemplary workloads include In-Memory Databases, scientific computation and multimedia applications. This project is a continuation of preceding projects conducted in the Spring and Fall Periods of 2018.

Researchers

Principle Investigator: Prof. Dr. Andreas Polze || Contact Author: Max Plauth || Hasso Plattner Institute, Germany

Abstract

Cardiovascular diseases are the leading cause of death globally. Telemedicine interventions were shown to reduce the percentage of days lost due to unplanned cardiovascular hospital admissions and all-cause mortality. This project will train an AI-System that could help to scale such telemedical interventions by preprocessing and prioritizing.

Researchers

Principle Investigator: Prof. Dr. Andreas Polze || Contact Author: Jossekin Beilharz || Hasso Plattner Institute, Germany

Abstract

During the course of the master seminar "Practical Introduction to Deep Learning for Computer Vision", students are learning how to apply state-of-the-art computer vision methods for the analysis of a large photo collection of the Getty Research Institute in Los Angeles. Students will learn how to deal with powerful GPU computer clusters.

Researchers

Principle Investigator: Prof. Dr. Christoph Meinel || Contact Author: Christian Bartz || Hasso Plattner Institute, Germany

Abstract

The project objectives are focused on research and development of methods, technologies, tools and software systems for efficient storage, processing, analysis, mining and visualization of Big mobility, transport and health-related data collected using mobile crowd sensing and Internet of Things paradigms in Smart Cities.

Researchers

Principle Investigator: Prof. Dragan Stojanovic || Contact Author: Prof. Dragan Stojanovic || University of Nis, Serbia

Abstract

Contact tracing is a key element in countering an epidemic. We propose a probabilistic network model that by exploiting interactions between individuals infers their probability of being infected. This will enhance the information available and will enable the design of more focused restriction and control policies and evaluate their effects.

Researchers

Principle Investigator: Elisabetta Fersini || Contact Author: Elisabetta Fersini || University of Milan - Bicocca, Italy

Abstract

Our preliminary work demonstrates the power of combining Raman spectroscopy and Deep Learning for a fast and non-invasive diagnosis of SARS-COV-2 infection from human salivary samples. In the next steps, we would like to investigate more complex DL architectures enabled by an increased data collection  to achieve proper clinical settings standards.

Researchers

Principle Investigator: Full Professor Vincenzina Messina || Contact Author: Dario Bertazioli || University of Milan - Bicocca, Italy

Abstract

The aim of this project is to check the scalability of parallel, network intensive microbenchmarks and application written in Java, using the PCJ library, HPC Challenge 2014 award-winning Java library for high-performance parallel computing, on the 1000 Core Cluster - with high performance Ethernet interfaces.

Researchers

Principle Investigator: Dr. Marek Nowicki || Contact Author: Dr. Marek Nowicki || Nicolaus Copernicus University in Toruń, Poland

Abstract

The WEEVIL6 project is designed as the extension of the previous one (WEEVILF), which was developed using the RX600S5-1 server from the HPI Future SOC Lab. This document details the main aims, scope and schedule development for the WEEVIL6 project.

Researchers

Principle Investigator: Prof. Dr. Carlos Juiz || Contact Author: Belen Bermejo || Universitat de les Illes Balears, Spain

Abstract

The goal of this project is to develop a non-intrusive automatic data collection mechanism to collect images of coral reefs in the Vamizi Island, in the north of Mozambique. Object detection algorithms will be used in real-time to automatically photograph, detect and classify fish and other marine species that will pass by the cameras.

Researchers

Principle Investigator: Erwan Sola || Contact Author: Luís Pina || Lúrio University, Mozambique

Abstract

The computer resources have helped us greatly to systematically investigate the influence of parameter settings on algorithm performance for problems with interconnected components. However, we have barely been able to scratch the surface to date, we are kindly asking for an extension of our access to your compute resources.

Researchers

Principle Investigator: Dr. Markus Wagner || Contact Author: Dr. Markus Wagner || University of Adelaide, Australia

Abstract

The Wildenstein Plattner Institute is undertaking a massive digitization project, with the goal to make millions of previously unpublished cultural heritage information available to the broader public. Because of the mass of information, it can only be processed using ML techniques like advanced OCR and image processing algorithms.

Researchers

Principle Investigator: Prof. Dr. Christoph Meinel || Contact Author: Christian Bartz || Wildenstein Plattner Institute Inc., United States