Modul
Practical Course: Advanced Topics in High Performance Computing, Data Management and Analytics [M-INFO-105870]
Credits
6Recurrence
Jedes SemesterDuration
1 SemesterLanguage
EnglishLevel
4Version
1Responsible
Organisation
- KIT-Fakultät für Informatik
Part of
Bricks
Identifier | Name | LP |
---|---|---|
T-INFO-111803 | Practical Course: Advanced Topics in High Performance Computing, Data Management and Analytics | 6 |
Competence Certificate
See partial achivements (Teilleitung)
Competence Goal
Students know and can apply tools and techniques in the fields of high-performance computing, data management and data analysis. They acquire the possibility to analyze complex scenarios and develop solutions for this. Besides working on the content, students improve their competences in communication and presentation.
Content
Participants will have the chance to deepen their knowledge of high-performance computing, data management and data analysis and to apply it in a practical way. The tasks to be worked on come from the subfields:
• HPC simulations (e.g., parallelization, MPI, performance engineering)
• HPC systems and operating environment (e.g., On Demand File Systems, Infiniband Networks, Job Scheduling)
• Machine Learning and Data Mining (e.g., RapidMiner, scikit)
• Data-Intensive Computing (e.g., Hadoop, Spark).
• HPC and data analysis with Python (e.g., Numpy, Scipy, Pandas, Dask, Parsl)
• Distributed & Parallel File Systems (e.g., glusterFS, BeeGFS)
• Object Storage (e.g., S3, CEPH)
• Data Management System (e.g., dCache, iRods)
• Databases (e.g., SQL, NoSQL)
• Workflow management systems for HPC and data analysis (e.g., FireWorks, AiiDA, SimStack)
• Opportunistic resource integration and utilization (e.g., using COBalD/TARDIS)
• Authentication and authorization infrastructure (e.g., OpenID, SAML)
Students are individually supervised by scientific staff of the Scientific Centre for Computing and can apply their skills in a practical and research-oriented way by being involved in current research tasks (e.g., Helmholtz program, BMBF and EU projects).
Recommendation
Knowledge in the area of databases, data management, data analytics, parallel computing is helpful.
Workload
3 SWS = 150 h per semester
• 12 h in meetings during the semester (kick-off, regular meetings with the supervisor, final meeting including presentation)
• 18 h preparation of meetings
• 120 h working on the topic and preparation of the exam