Assistance can be provided to choose an appropriate platform for research projects, improve the workflows, and adapt them to new challenges, regardless of where these computations are being run – on personal laptops, terminal servers or the HPC cluster.
Read up on the scientific computing services to find out how to improve scientific computing workflows with iDiv’s support!
Christian Krause
Technical Employee (HPC Cluster)
Phone: +49 341 9733144
Email: christian.krause@idiv.de
Platforms and Resources
RStudio Servers
The RStudio servers provide access to the RStudio Web IDE (Integrated Development Environment). No installation is required, and the servers can be accessed from anywhere in the world. This platform is suitable for interactive work and medium-sized computations.
Terminal Servers
The MS Windows Terminal Servers provide access to applications that run only on Windows, e.g. ArcGIS. Login is possible via RDP (Remote Desktop Protocol). This platform is suitable for interactive work and medium-sized computations.
High-Performance Computing Cluster
The High-Performance Computing (HPC) cluster provides access to virtually all Linux software. High volumes of jobs can be submitted to the cluster. The cluster scheduler automatically assigns resources to these jobs and processes them over time – no interaction is required. This platform allows for the highest degree of automation and is suitable for the largest computations in processing power, memory, disk storage and time.
Resources
| RStudio Servers | Terminal Servers | HPC Cluster | |
|---|---|---|---|
| processors | ~ 24 | ~ 50 | ~ 2,300 |
| memory | ~ 280GB | ~ 600GB | ~ 27TB |
| disk storage | ~ 5TB | ~ 35TB | ~ 2.5 Petabyte |
SCaaS – Scientific Computing as a Service
Service Steps
1) Consultation
- Requirement analysis
- Sketching workflows
2) Development
- Developing tools
- Implementing workflows
- Integrating data management
3) Maintenance
- Providing an issue tracker for ongoing development
- Teaching how to maintain and adapt the tools
Seminars
The basics of scientific computing are taught in seminars. The acquired knowledge can be advanced as part of the hands-on sessions (see below).
Topics
Possible topics include but are not restricted to:
- Introduction to High-Performance Computing
What is high-performance computing? - Version Control and Collaboration with git
How to manage source code, scripts and papers with the version control system git and how to orchestrate collaboration beyond using comments in MS Word. - Linux, the Command Line and Shell Scripting
How to use Linux via its flexible and powerful user interface, the command line. - Reproducible Research
How to create workflows with baked-in reproducibility.
Sessions
Hands-on sessions are offered to deal with specific topics in depth. The sessions are ideally suited to find solutions to particular problems or to gain knowledge in a specific area of scientific computing.
Topics
Possible session topics include but are not limited to:
- Workflow management
Pipelines
Portability
Automatic result validation / verification
Reproducible research - Data management
Acquisition
Archival - Project administration
Git
Testing - Performance analysis
Monitoring
Debugging - Software design and development
Optimisation
Parallelisation