High-Performance Computing (HPC) provides researchers with the ability to expand their data processing, simulation and computation across hundreds of cores. This has led to widespread use across many disciplines.
ITS Research support, and provide access to Apocrita, a 300+ node heterogeneous HPC Cluster with over 5,500 compute cores, located in the Jisc Shared data centre in Slough. The cluster has many applications installed ready for you to use; core applications (including Stata, Gaussian and MATLAB); programming languages (e.g. C, C++, Fortran, R, Python); and domain-specific research applications (e.g. for bioinformatics and computational chemistry).
Why use Apocrita?
Using an HPC cluster can substantially reduce the time taken to compute problems by running analyses in parallel or breaking them up into pieces. Both strategies allow for the problem to be worked on by many separate compute cores (CPUs) at the same time. Any Researcher who commonly finds themselves regularly performing computational work on their machine should consider moving their workflow to the HPC Cluster, where you can significantly increase analysis throughput, or enable computations which are not feasible on a standard desktop machine while utilising a scalable, backed-up storage system to store your research data.
The Apocrita system has been specifically designed to meet the needs of a wide variety of research in the areas of our three faculties, including GPU nodes to facilitate Machine Learning workloads, low latency Infiniband networking for parallel jobs favoured by astronomy and computational chemistry, plus high-performance storage to aid genomic analyses.
Please visit our HPC website: docs.hpc.qmul.ac.uk - for more information, details on applications available on the cluster, or to request an account.
Applications and Language
ITS Research provides a wide range of commercial and open-source applications on the HPC cluster.
Popular applications include:
- Gaussian - QMUL has licences for the Gaussian computational chemistry application, allowing QMUL researchers to use it locally on their computers and the HPC cluster;
- MATLAB - This is an environment for numerical computation used in science and engineering, and QMUL has a site licence allowing QMUL researchers to use it locally on their computers and the HPC cluster;
- Stata - This is a statistics package used by a broad range of researchers. Licences for Stata/MP16 are available for 11 concurrent users on the QMUL Apocrita HPC cluster;
- R - is a language and environment for statistical computing and graphics;
- TensorFlow - is an open-source library for machine learning, often utilising GPUs to accelerate performance;
- Python - is an interpreted high-level programming language for general-purpose programming;
- Ansys - Engineering application is licensed for 128 cores on Apocrita;
- Anaconda - is a distribution of the Python and R programming languages for data science and machine learning applications. It includes more than 250 popular data-science package;
To see a more detailed list of HPC applications we provide please visit: docs.hpc.qmul.ac.uk/apps
You can also view the licensing arrangements for other relevant software by searching our A-Z list of services.
Getting started on the Cluster
New cluster accounts for eligible users can be requested here, there are also sections on how to log in, change your password, submit a job, and a description of the various nodes available on the cluster. Support hours for High-Performance Computing are 9 am to 5 pm weekdays, and you can contact the team at: firstname.lastname@example.org.
Use of Apocrita is free, but it is a shared service. If a research project is going to make intensive use of the cluster then we request that you add compute capacity as part of your grant application. This allows the cluster to be expanded and to improve the level of service available. This will also provide you with a dedicated computational resource for your research while leveraging the high-performance infrastructure.
Information on the different node types available on Apocrita can be found here: docs.hpc.qmul.ac.uk/nodes/
If you need to discuss your requirements further, please contact the ITS Research Consultants: email@example.com.
GitHubIf you intend to write code then we recommend using a version control tool such as git. GitHub is a web-based Git repository hosting service, which offers all of the distributed revision control and source code management (SCM) functionality of Git as well as adding its features. Unlike git, which is strictly a command-line tool, GitHub provides an additional web-based graphical interface and desktop as well as mobile integration. It also provides access control and several collaboration features such as code reviews and integration with test suites.
QMUL host a GitHub Enterprise instance for additional control and security. All Apocrita HPC users are automatically registered to use the QMUL GitHub service. Alternatively, you can request a GitHub only account.
If you run an undergraduate course module which encourages the use of GitHub for version control and/or collaboration, you can have oversight of the activities via a GitHub organisation. Please get in touch with us to arrange this, including bulk addition of accounts for students.
Queen Mary University of London has an active Globus license for transferring, sharing, and discovering data via a web browser. It allows collaboration with other researchers, without requiring them to have accounts on QMUL systems. Globus uses widely-adopted industry standards for authentication and authorisation, as well as GridFTP and HTTPS protocols. Globus lets you share data with collaborators at other research institutions, whether your data exists on Apocrita, Tape archive, Scientific instrument or Laptop /PC.
If you would like to use Globus on Apocrita, please contact the ITS Research Consultants: firstname.lastname@example.org.
Software Development Support
Our Research Software Engineers can assist with reviewing and improving your research software, via writing/modifying code, compiler optimisation, or discussion about good practice. They have a particular focus on Fortran and parallel computing, and would particularly welcome enquiries from anyone requiring assistance in this area.
Research Computing Support, Education and Training
Our training site provides primers on getting started on the cluster, short training sessions on basic Linux usage. We can also present face-to-face training sessions on these subjects and can arrange training sessions from software vendors if there is sufficient demand.
Specialised Teaching Support
For the interface between teaching and research, we can provide systems and programming support, and have also assisted with using the HPC cluster in undergraduate teaching labs for Computational Fluid Dynamics, and the Genomics MSc. For enquires please email ITS Research Consultants: email@example.com to arrange a chat.
Research Data Storage
It is now more important than ever to ensure that your data is stored in a resilient and replicated fashion. Of course, none of this is of any use unless it is easily accessible by you the researcher as part of your normal workflow. We provide a fully supported HPC storage solution, including the ability to easily increase storage capacity as required. Facilities for sharing data with collaborators (both within QMUL and externally) are also available. Additional research data management tools will be made available as required by the QMUL Research Data Management Policy.
Sometimes your research requirements will be so specialised that other offerings cannot meet your needs. In this instance, we can recommend suitable hardware and provide a managed hosting service.