Isambard-AI and Dawn AIRR supercomputers: Gateway route

- Funders: UK Research and Innovation, Arts and Humanities Research Council (AHRC), Biotechnology and Biological Sciences Research Council (BBSRC), Economic and Social Research Council (ESRC), Engineering and Physical Sciences Research Council (EPSRC), Innovate UK, Medical Research Council (MRC), Natural Environment Research Council (NERC), Science and Technology Facilities Council (STFC) - Funding type: Other

Last updated: 26 February 2026 - see all updates

Start application

This access route is open to UK-based researchers from academia, industry, and other organisations.

It offers flexible computational support for artificial intelligence (AI) related research and development projects, such as:

- first-time users of AI supercomputing resources

- testing of (novel) algorithms, code, and workflows

- benchmarking of algorithms, code, and workflows before applying for larger AI Research Resource (AIRR) opportunities

Each successful project will be awarded up to 10,000 graphics processing unit (GPU) hours on the Isambard-AI and Dawn supercomputers, to be used within three months of project start. No funding is provided.

Who can apply

This opportunity is open to eligible researchers from across the UK Research and Innovation (UKRI) remit.

To be a project lead, you must have a contract (of longer duration than your proposed project) with your organisation.

Your organisation must be one of the following:

- UK registered business of any size, with a Companies House registration number

- a research organisation eligible for UKRI funding

- charity

- not for profit

More information on the different types of organisation can be found in our funding rules and research organisations eligible for UKRI funding.

There is no limit to the number of applications from any one organisation.

There is no limit to the number of applications from any one researcher, however each application from that researcher must be standalone.

We welcome collaborative projects, however this route is not aimed at large research consortia.

We consider research technical professionals, including research software engineers, as academic employees. They are eligible to be a project lead or co-lead under the same terms as traditional researchers.

Equality, diversity and inclusion

We are committed to achieving equality of opportunity for all applicants. We welcome and encourage applications from people of all backgrounds and are committed to making our application process accessible to everyone.

What we're looking for

The Gateway route is an open and flexible pathway to request computational resources for users across academia, commercial and individual research. Users should be looking to explore novel and developing areas of AI research, as well as supporting upskilling across a wide range of AI-relevant experience.

Potential AI-related research topics

These include, but are not limited to:

Development of novel algorithms and software tools

Examples include:

- enhanced performance and efficiency: improved accuracy, processing speed and resource efficiency

- scalability and adaptability: developing algorithms that are more adept at handling big data, and that can enable AI models to apply knowledge from one domain to another

- robustness and reliability: developing algorithms that are resistant to adversarial attacks and that can provide clear explanations for their decisions

- innovation and new applications: novel creative AI algorithms can drive advancements in generative models, or that can that enhance human-AI collaboration to leverage the strengths of both humans and machines

- ethical and responsible AI: developing algorithms that detect and mitigate biases, and that ensure data privacy

Exploring AI-assisted workflow

Examples include:

Optimisation of experimental parameters

AI can be used to fine-tune experimental setups by analysing vast amounts of data to identify optimal conditions. This can significantly speed up research and development processes in fields like materials science and drug discovery.

AI-enhanced simulations

AI can enhance the accuracy and efficiency of simulations in various domains, such as climate modelling, financial forecasting, and engineering design. By integrating AI, simulations can become more predictive and less computationally intensive

Hypothesis generation

AI can assist researchers in generating new hypotheses by identifying patterns and correlations in data that may not be immediately apparent to humans. This can lead to novel insights and directions for research

AI-driven data collection, production, and synthesis

Examples include:

- data collection: AI can automate and improve the processes of data collection, ensuring higher quality and more comprehensive datasets

- data-set curation: using AI to help automate the curation and cleaning process, identifying and correcting errors, filling in missing values, and standardising data formats

- data-set creation: generating synthetic data or compiling existing data into a structured format

Within these novel and developing areas, we are particularly keen to hear from projects that contribute to delivering against the government’s six missions:

- strong foundations

- growing the economy

- an NHS fit for the future

- safer streets

- opportunity for all

- making Britain a clean energy superpower

And the modern Industrial Strategy’s eight sectors:

- advanced manufacturing

- creative industries

- life sciences

- clean energy

- defence

- digital and technologies

- professional and business services

- financial services

This route covers activities that fall into the following categories:

- fundamental research

- feasibility studies

- industrial research

- experimental development

See more on our categories of research and development.

Resources available

Each application can request 10,000 GPU hours on either of the AIRR services, Dawn or Isambard-AI.

The GPU hours must be used within three months of the project being awarded.

Projects may specify a preferred start date up to one month after submitting their application. While we will make every effort to honour this request, compute resources will be made available immediately upon project approval.

The project start date will not be able to be deferred once the award is made. Project extensions can only be requested on a case-by-case basis and under exceptional circumstances.

This application process is purely for compute resource. No funding is available to successful applicants.

Trusted research and innovation

UK Research and Innovation (UKRI) is committed in ensuring that effective collaboration in research and innovation takes place with integrity and within strong ethical frameworks.

Trusted research and innovation (TR&I) is a UKRI work programme designed to help protect all those working in our thriving and collaborative international sector by enabling partnerships to be as open as possible, and as secure as necessary.

Our TR&I principles set out UKRI’s expectations of projects awarded through this route in relation to due diligence for UK and international collaboration.

Subsidy control and State aid where applicable

All AIRR access routes provide awards in line with the Subsidy Control Act 2022.

In the ‘Lead organisation details’ section of your application you will be asked questions to indicate if State Aid or Subsidy applies to your organisation, including your subsidy history.

If you are unsure about your obligations under the Subsidy Control Act 2022 or the State aid rules, you should take independent legal advice. We are unable to advise on individual eligibility or legal obligations.

You must always make sure that the resources awarded to you are compliant with all current Subsidy Control legislation applicable in the UK.

Project finances

Project applications from businesses, or non-academic organisations undertaking feasibility studies, industrial research or experimental development projects, will need to state all eligible project costs. Academic organisations do not need to provide this information. The information on project costs is required to ensure awards are made inline with the Subsidy Control Act 2022.

Remember that no funding is available through this route, only compute resource.

For an overview on what costs you can include in your project, see Costs guidance for non-academic organisations. Note this is general guidance. You can also view our application finances video.

The AIRR programme would welcome the following project costs as aligning to the policy objectives:

- software development

- staff training related to AI

Support ratios

Project applications from businesses, or non-academic organisations undertaking feasibility studies, industrial research or experimental development projects, will need to ensure that the balance between your total eligible project costs and the amount of compute resources awarded is funded by your organisation. Academic organisations are not subject to this requirement.

For feasibility studies and industrial research, the maximum percentage of the overall eligible project costs that can be compute resources awarded through this route are:

- up to 70% if you are a micro or small organisation

- up to 60% if you are a medium-sized organisation

- up to 50% if you are a large organisation.

For experimental development, the maximum percentage of the overall eligible project costs that can be compute resources awarded through this route are:

- up to 45% if you are a micro or small organisation

- up to 35% if you are a medium-sized organisation

- up to 25% if you are a large organisation

For more information on company sizes, refer to the company accounts guidance.

If you are applying for an award funded under State aid Regulations, the definitions are set out in the European Commission Recommendation of 6 May 2003.

How to apply

You can use the AIRRPortal to apply to use AIRR.

Apply to use AIRR

General guidance on using the AIRRPortal can be found at:

AIRRPortal: how to apply (UKRI guidance)

You must read this guidance fully before you start your application. If your application does not follow this guidance, it may be rejected.

Before submitting, it is the lead applicant’s responsibility to ensure that:

- all information provided in the application is accurate

- all project information details requested have been provided

- the application meets the eligibility and scope criteria for the chosen access route

- the application contains the additional documents requested in the guidance (templates for the documents to be uploaded are available on the AIRRPortal)

- applications use official organisational email addresses that are unique to each team member and clearly identify them by name (for example, applications must not include Gmail addresses or corporate functional IDs)

Do not include links to external web resources. Applications containing such links will be deemed ineligible.

You may use generative AI to help prepare your application. However, exercise caution and ensure the content remains accurate and appropriate. For more details, refer to the UKRI policy on the use of generative AI in application and assessment.

Application structure

The application consists of the following components:

AIRRPortal online form, with three sections (not scored by assessors):

- project details (which includes the ‘Upload supporting documentation’ option, where the attachments requested below are to be uploaded)

- resource requests (AIRR service requested)

- project team membership

Supporting documentation to be uploaded (templates for both attachments are available on the AIRRPortal):

- further project and organisation details, including compliance and project costs and subsidy history (not scored by assessors)

- assessment questions (scored by assessors)

Processing personal data and data sharing

UK Research and Innovation (UKRI) will need to collect some personal information to manage your application.

We will handle personal data in line with UK data protection legislation and manage it securely. For more information, including how to exercise your rights, read our privacy notice.

UKRI will need to share the application and any personal information that it contains with the Department for Science, Innovation and Technology (DSIT) so that they can participate in the assessment process.

See more information on how DSIT uses personal information.

How we will assess your application

We reserve the right to modify the assessment process as needed.

Assessment

Gateway applications will be subject to a light touch review by a member of the delivery team. The UKRI and DSIT delivery team will use the following assessment criteria:

- does the project clearly align with the scope of the AIRR programme, with a well‑defined explanation of the novelty or innovation in its use of AI?

- is there a clear description of the project’s anticipated outputs and how these outputs will benefit the project team and contribute to wider research or innovation activities?

- is there a clear explanation of how the project supports the government’s modern industrial strategy, missions, or other relevant policy priorities?

Awarding resources

The aim is that GPUh resources will be awarded at fortnightly batching dates, with an estimated time-to-resource-access for successful applications of one month after the date of submission.

Resources available for new projects at any fortnightly batching date will be limited, based on the current live portfolio of Gateway awards and the overall usage levels of the AIRR services.

The following process will be used to assign any available resources.

Following the assessment, applications will be allocated to one of three tiers, based on the above assessment.

Resources will be allocated to the applications in the top tier initially, using partial randomisation as required, with priority given to applicants who have not used AIRR services previously.

Applications in the middle tier may be recommended for resources using partial randomisation.

Resources are allocated until the budget is exhausted throughout the top two tiers.

Unsuccessful applications will not be automatically reconsidered at future batching dates.

The UKRI or DSIT delivery team will make the final allocation decision.

What happens if you receive an award offer?

If your project application is successful, UKRI will send the project lead a formal offer letter, which will contain the award terms and conditions, any subsidy notice required and a link to the online equality monitoring form.

When the equality monitoring form has been completed, the project lead will be emailed a project link from the AIRRPortal.

By logging in to the AIRRPortal and accessing the compute resources for your project, you will have accepted the UKRI terms and conditions.

Top-up policy

You can request a one-off, additional 50% to your original award (subject to a subsidy re-assessment) if you have used over 90% of your original allocation and have sufficient time left on the project to use the additional GPUh.

Additional info

Background

The AIRR programme intends to address the significant shortage of publicly available computing resources in the UK. In January 2025, the government announced expanding AIRR capacity, by at least 20 times by 2030, as part of the AI Opportunities Action Plan.

The government has committed to spending an extra £1 billion to scale up our compute power by a factor of 20, giving Britain the power to become an AI leader.

AIRR is a partnership between:

- UK government

- UK Research and Innovation (UKRI)

- University of Cambridge

- University of Bristol

- Hewlett Packard Enterprise (HPE)

- Nvidia

- Intel

- Dell

AIRR compute clusters

The government is investing significantly in the Isambard-AI and Dawn AIRR clusters and will have invested over £350 million by 2030.

Isambard-AI (University of Bristol)

The Isambard-AI facility is the UK’s most powerful public compute facility. It is made up of 5,448 Nvidia GH200 Grace-Hopper superchips (supplied by HPE) and operated by the University of Bristol at Bristol’s National Composites Centre.

See further information on the Isambard-AI system.

Dawn (University of Cambridge)

The Cambridge Dawn facility is made up of 1,024 Intel Data Centre GPU Max 1550 GPUs. It is a close partnership between Cambridge, Intel and Dell at the Cambridge Open Zettascale Lab.

See further information on the Dawn system.

Supporting documents

AIRRPortal: how to apply (UKRI guidance)