The ARC team provide and support a comprehensive Advanced Research Computing (ARC) support service for researchers in all faculties at the University of Leeds.
The service consists of a large High Performance Computing (HPC or Supercomputing) resource consisting of several clusters which are used for all forms of predictive modelling, data analysis and simulation which are too large, too complex or too time consuming to be performed on a Desktop PC. At the current time, researchers have access to three local HPC clusters (ARC2, ARC3 and MARC1) and an additional two machines (Polaris and HERC1) as part of the N8 HPC research consortium. In total, these provide over 15000 computational cores and over 1PB of data storage.
In addition, the team can support researchers in the use of Cloud services (such as those from Amazon and Microsoft), ‘Big Data’ technologies such as Apache Spark & NoSQL databases and GPU Computing.
We also provide a range of consultancy and training services including:
* user guidance and training
* installing and supporting software applications on the HPC service
* improving software performance and handling large data sets
* guidance and support in meeting current and future research challenges with HPC
* how to incorporate HPC costings into grant applications
Because of the way the service is funded, there is no charge to research projects or to research groups and users of all skill levels and from all faculties can use the facilities. We can support problems of any size in any research area and as long as there’s a Linux version of your application, we can probably make it work.
The team consists of Research Computing Consultants Martin Callaghan and Joanna Leng who support and train research users in a range of computational and High Performance Computing techniques (everything through automating workflows with Shell scripts, through Python and R scripting, profiling and debugging codes, parallel programming and visualisation) and HPC System Administrators Mark Dixon and Mike Wallis who design and manage the HPC systems and provide consultancy and support to users.