If a researcher has a specific problem or requirement using BEAR compute services, such as BlueBEAR, BEAR Cloud VM, or CaStLeS Compute, then we can help. A researcher can request help with any aspect of using the range of BEAR compute services by filling in the Help with BEAR Compute request.
We maintain an extensive range of software on the BEAR systems. If a researcher has need of specific software to be used on either BlueBEAR, BEAR Cloud VM, or CaStLeS Compute then, where possible, we can install it for you. To ask for new software to be installed on BEAR, please make a New BEAR Software Request.
Example tasks might include debugging or profiling research software, or advising how to optimise batch compute jobs.
When a researcher has issues running third party software then BEAR Software RSEs can work with the researcher to find a solution.
A researcher may have existing sequential code and we can offer advice on the possible methods for parallelisation (to multi-core, multi-node, or GPU systems), which will enable better utilisation of the computing resources available to the researcher.
BEAR Software RSEs can advise researchers on how to apply for access to, and facilitate the use of, further computing resources available at the regional (HPC Midlands+), national (ARCHER), and international level (Prace).
BEAR Software RSEs may seek to initiate conversations with researchers if we see an opportunity to find out more about interesting techniques or to get more out of the computing resources.
The Research Software Group, in conjunction with Advanced Research Computing, provide training courses and workshops to researchers at the University. These include an Introduction to BEAR Linux and our HPC system (BlueBEAR); hands-on workshops on the foundations of coding in either Python or R; and workshops on AI and Deep Learning.