Compute Services
These shared IT platforms provide the processing power needed to run applications, analyse data, and support teaching, research, and operational work without individuals needing to manage their own servers.
People at the University use compute services when their work goes beyond a standard laptop or desktop - for example, handling large datasets, running specialist teaching software, developing or testing systems, or carrying out compute‑intensive research - with the benefit of scalable, secure, and centrally supported infrastructure.
Our Services
Select the tabs below to learn more about the services we can support you with.
Description and Features
Research IT provides access to a range of High Performance Computing (HPC) facilities to support AI, numerical computation, and data-intensive research workloads. Three key systems are available:
- ISCA – the University's onsite HPC cluster, available to all of Exeter's research and education communities for CPU-intensive tasks, with GPU capabilities being progressively introduced throughout 2026.
- Athena – a smaller onsite HPC system serving the Cornwall campus community.
- Isambard-3 – access to this GW4-alliance supercomputer is available through a standing University allocation, supporting more demanding computational workloads.
Research IT staff provide guidance and advice on selecting the most appropriate system for your work and can assist with onboarding to both University-managed and external national HPC services.
Support and availability
The current offerings on available sysstems:
- ISCA / Athena: CPU compute is available, with best efforts GPU capabilities being introduced throughout 2026.
- Isambard 3: CPU compute is available from a standing share provided through the GW4 alliance.
Support is provided free of charge for research use of the systems, but not at the level of specific software support (please refer to Research Enablement Services). Support is available during business hours. The platforms are monitored and operated 24/7/365.
Please note that Isambard-AI access is managed through AIRR resource requests and not by the University of Exeter.
Requesting the service
Internal services are free at the point of use, as is the standing allocation for Isambard. Access requests are made via the Research IT and Research Software and Analytics teams. Please raise an internal helpdesk ticket for access.
Who we work with
Research Software and Analytics (RSA) BRiCS
Links
- ISCA Documentation INTERNAL
- Athena Documentation INTERNAL
- Isambard-3 Exeter Documentation INTERNAL
- Isambard-3 Documentation
- Isambard-3 Portal
Description and Features
Research IT supports the design, deployment, and ongoing management of both on-premise and cloud-based compute infrastructure for research projects. Whether you need a single virtual machine or a complex multi-component environment, the team can help.
Supported platforms include OpenStack for on-premise virtualisation, and leading cloud providers Microsoft Azure and AWS for scalable, flexible cloud deployments. The service covers:
- Deployment of virtualised, and containerised infrastructure.
- Configuration and ongoing system administration.
- Guidance on the most appropriate hosting model for your research requirements, budget, and data sensitivity.
Support and availability
Please raise a ticket with Research IT through the Service Desk to discuss your infrastructure requirements and available support arrangements.
Description and Features
Research IT can help you design resilient, scalable, and fit-for-purpose digital research infrastructure. Working with other IT teams as required, we can provide architectural design services tailored to the needs of research-oriented systems, including advice on system patterns, component selection, and integration approaches.
The service offers:
- Standard architecture templates and proven infrastructure patterns for common research system types.
- Bespoke architectural design for novel or complex requirements.
- Guidance on how proposed designs integrate with the University's existing hosting and compute capabilities.
Early engagement with Research IT at the design stage helps ensure that infrastructure is both technically sound and appropriately costed for funding applications.
Support and availability
Please raise a ticket with Research IT through the Service Desk to begin an infrastructure design consultation.
Description and Features
Research IT provides interactive, on-demand compute environments designed to support teaching and education activities involving data analysis and programming. These platforms give students and educators access to familiar analytical tools without the need to manage local software installations.
Available environments include:
- Temporary virtual machines for isolated testing and research processing tasks, allowing users to create and configure their own environments. •Cloud-hosted virtual machines providing scalable, fully managed compute environments for teaching and research.
- RStudio/Posit servers providing a web-based interface for coding and analysis in R — four dedicated servers are available (two teaching-specific, two education-specific), licensed for teaching and education use.
Note: RStudio servers are licensed for teaching and education purposes only and are not intended for active research use.
Support and availability
Please raise a ticket with Research IT through the Service Desk for support with existing environments or for guidance on creating new environments.
Description and Features
In addition to University-managed systems, Research IT can support researchers in accessing a broad range of national and external high-performance and specialist compute resources. These services can provide significant additional capacity for large-scale or specialist workloads that exceed local capabilities.
National resources the team can advise on and assist with include:
- ARCHER2 – the UK national supercomputer for academic research.
- Isambard-3 – GW4-alliance Tier 2 HPC system.
- DiRAC – distributed research infrastructure for theoretical astrophysics and particle physics.
- Cirrus – EPSRC-funded HPC service.
- Isambard-AI – national AI-focused compute resource (access via AIRR applications).
- JASMIN – specialist platform for environmental and earth science data analysis.
Research IT can advise on eligibility, application processes, and how to make best use of these resources.
Who we work with
Research Software and Analytics (RSA)
Links
Description and Features
As AI-driven research workflows become increasingly common, Research IT provides advisory support to help researchers design, configure, and run AI pipelines effectively on available compute infrastructure.
The service covers:
- Advising on the design of AI and machine learning workloads to make best use of available GPU and CPU resources.
- Guidance on data ingress — how to efficiently move datasets into compute environments for AI processing.
- Support with egressing and managing results data from AI pipelines.
- Advice on integrating AI workloads with the University's HPC systems and external national compute resources.
This service is particularly valuable for researchers new to large-scale AI computation, or those looking to optimise existing pipelines.
What we offer
Please contact Research IT to arrange an AI workload advisory consultation.
Who we work with
Research Software and Analytics
More information
https://www.exeter.ac.uk/about/strategies/enabling-ai/#catalogue