Im Folgenden finden Sie eine Aufstellung der zur Verfügung stehenden Themen. Die angegebene Literatur versteht
sich als Startlektüre und weitere Literatur sollte selbstständig recherchiert wertden.
The suggested topics are listed below. The literature indicated here is intended as a starting point only - further literature should be researched independently.
1) Intel's Global Extensible Open Power Manager (GEOPM)
A framework for power/energy optimizations targeting
High Performance Computing (HPC) by dynamic coordination of hardware settings across system compute nodes utilized
by a given parallel application in response to the application's behavior and requests from the resource management
and scheduling system.
2) Parallel and Distributed Deep Learning
Deep Neural Networks (DNNs) are becoming an important vehicle for
modern large-scale computing applications, while their training in most cases still requires a significant amount of
time. The topic discusses the training problem and describes possible approaches for its parallelization.
3) Tensor Processing Units (TPUs) and their In-Datacenter Deployment Analysis
A Tensor Processing Unit (TPU)
is application-specific integrated circuit (ASIC) developed by Google for accelerating the inference phase of
artificial neural networks. The topic describes these AI accelerators and discusses their indatacenter
- Jouppi, Norman P., Cliff Young, Nishant Patil, David Patterson, Gaurav Agrawal, Raminder Bajwa, Sarah Bates et al. "In-datacenter performance analysis of a tensor processing unit." In Computer Architecture (ISCA), 2017 ACM/IEEE 44th Annual International Symposium on, pp. 1-12. IEEE, 2017.
4) Container technologies (e.g. Singularity, Docker, Charliecloud, Shifter)
Containers, being instances of
an Operating System (OS) level virtualization, are getting more appealing due to their higher efficiency as compared
to the full, hardware-level, virtualization. The topic describes the concept of the containers, and discusses the
differences among main software containers like Singularity, Docker, Kubernetes, Charliecloud, and Shifter that are
currently used in HPC.
- Abdelbaky, Moustafa, Javier Diaz-Montes, Manish Parashar, Merve Unuvar, and Malgorzata Steinder. "Docker containers across multiple clouds and data centers." In Utility and Cloud Computing (UCC), 2015 IEEE/ACM 8th International Conference on, pp. 368-371. IEEE, 2015.
- Kurtzer, Gregory M., Vanessa Sochat, and Michael W. Bauer. "Singularity: Scientific containers for mobility of compute." PloS one 12, no. 5 (2017): e0177459.
5) Deep Neural Networks for Malware Detection in Executables
As the use of computing devices increases in
every day life, malware detection continues to remain a serious challenge for corporations, governmental agencies,
and individuals. Today malware detection systems still heavily rely on heuristic and signature-based methods, where
signature represents set of rules that are generally specific and thus usually fail to capture a new malware. This
topic discusses an alternative approach, that uses neural networks and the raw bytes of the binary program itself to
determine maliciousness without executing the target application.
- David, Omid E., and Nathan S. Netanyahu. "Deepsign: Deep learning for automatic malware signature generation and classification." In Neural Networks (IJCNN), 2015 International Joint Conference on, pp. 1-8. IEEE, 2015.
- Raff, Edward, Jon Barker, Jared Sylvester, Robert Brandon, Bryan Catanzaro, and Charles K. Nicholas. "Malware detection by eating a whole exe."
6) Explainable System Predictive Maintenance
This topic discusses various Machine Learning based approaches aimed at detection of anomalies within HPC system and surrounding ecosystem. In doing so, a special focus will be attributed to the explainability of the discussed predictive modeling solutions.
7) Intel oneAPI
Current HPC workloads are becomming more data-driven. This pushes further the need for diverse computer architectures (e.g. CPU, GPU, FPGA, etc.) in order to deliver the required compute performance for the various specialized data-centric workloads. However, taking advantage of multiple HW architectures is a challenge for developers, as it is complex and time-intensive due to the lack of code-reuse possibilities; appropriate programming models that do not degrade the application performance; etc. Intel's oneAPI is a single unified programming model that aims to ease the development accross multiple architectures. This topic introduces the basics oneAPI programming model and discusses its applicability across various HW architectures.
8) Contemporary HPC and the World's Fastest Supercomputer Summit
The topic discusses the supercomputing
today by analyzing current TOP500 list and identifying the major trends. Additionally, it describes Summit system,
deployed at Oak Ridge National Laboratory and recognized as the world's fastest computer back in June' 18 during
International Supercomputing Conference by delivering 122.3 PFLOPS performance during the LINPACK run. Since then
the system has been upgraded to deliver 143.5 PFLOPS LINPACK performance making it still the fastest supercomputer
in the world (TOP500 November'19 rankings). The topic discusses the system design aspects and describes its compute
and cooling infrastructure.
RISC-V (pronounced as Risk-Five) is a free and open reduced instruction set computer (RISC) Instruction Set Architecture (ISA). Originated in UC Berkeley back in 2010, RISC-V is continuously gaining its popularity due to its open and modular design. This topic provides introduces and provides a high-level overview of the RISC-V standard.
The topic discusses the open-source container orchestration system, Kubernetes, that assists developers in the deployement, scaling, and management of their large-scale applications.
11) Introduction to Quantum Computing
The topic introduces the basics of qunatum computing and discusses its current state, the applicability range, challenges, and perspectives.
12) Machine Learning Based Classification Over Encrypted Data
The HPC systems today enable the application
of Machine Learning (ML) based methods in various domains, ranging from face recognition to medical or genomics
predictions, as they provide the required data storage and computational power. However, many of these ML based
applications rely on sensitive data usage making it important to control the privacy of the considered data and the
underlying classifier. The topic discusses ML based classification techniques over encrypted data.
13) Cross-Architectural Modelling of Power Consumption Using Neural Networks
Energy consumption is becoming
a dominating factor for the Total Cost of Ownership of many supercomputers, making it important to keep energy costs
in budget and to operate within available capacities of power distribution and cooling systems. The topic
considersprediction of power consumption of HPC systems utilizing artificial neural networks that use data
obtained from hardware performance counters. The topic discusses the accuracy of the proposed, portable across
different micro-architectureimplementations, methodology and outlines the advantages against its simpler,
linear-regression based, counterparts.
14) Next generation arithmetic for HPC and AI
Posit arithmetic, a form of universal number (unum)
computer arithmetic, is designed as a direct drop-in replacement for IEEE Standard 754 floating-point numbers
(floats), provide compelling advantages over floats, including larger dynamic range, higher accuracy, better
closure, bitwise identical results across systems, simpler hardware, and simpler exception handling.Posits never
overflow to infinity or underflow to zero, and 'Nota-Number' (NaN) indicates an action instead of a bit pattern. The
topic discusses the posit arithmetic and outlines its advantages against fixed-pointarithmetic approaches currently
used for AI and signal processing.
15) Explainable Aritficial Intelligence (XAI): Opportunities and Challenges
Recent years have witnessed the extension of applicatibility range of artificial intelligence (AI) to different domains. In fact, AI is already assisting humanity in building better security systems, efficient transportation and web searches, personalized advertising, etc. Sometimes, it even decides whether a given person should be approved for a bank loan. However, the lack of complete explainability of various approaches serves as an obstacle for the further deployement of the proposed AI-based solutions. This topic will discuss the concepts of explainable AI (XAI), and outline the existing opportunities and challenges.
16) AI for Detecting Breast Cancer: explainability and accuracy
The topic reviews the state-of-the-art Artificial Intelligence (AI) based solutions for breast cancer detection, dicusses their explainability and outlines the currently existing limitations.
- Lamy, Jean-Baptiste, Boomadevi Sekar, Gilles Guezennec, Jacques Bouaud, and Brigitte Seroussi. "Explainable artificial intelligence for breast cancer: A visual case-based reasoning approach."
- Le, E. P. V., Y. Wang, Yuan Huang, S. Hickman, and F. J. Gilbert. "Artificial intelligence in breast imaging."
Nowdays, when checking emails, when purchasing from an online store, when reading a document, we always expect to have a time-efficient search engine working for us flawlessly in the background. Moreover, we typically expect these engines to be intelligent enough to support us with various manipulations with our (pending) search results (e.g. sorting, suggesting, etc.).
Elasticsearch is an open-source search engine library developed in Java. This topic introduces the basic concepts and principles of Elasticsearch, and outlines how typical search problems can be tackled with the help of Elasticsearch.
In the current era of big data, where understanding of various complex systems continuously pushes the need for the generation of even larger data volumes, makes the efficient data processing and storage an imperative and natural requirement.
This topic discusses the Hadoop platform that allows for distributed storage and data processing. The topic introduces Hadoop architecture, provides a high-level overview of its functionalities, and outlines its limitations.