This funding opportunity is to fund a range of high end computing (HEC) consortia, spanning EPSRC’s remit.
This is for communities that currently gain significant benefit, or are seeing rapidly increasing benefit, from the use of high performance computers (HPC) for modelling and simulation.
An HEC consortium is defined by EPSRC to be a network of computational researchers who are provided with a portion of EPSRC’s HPC resource, which it distributes among its members.
Each HEC consortium must be associated with a different research area within EPSRC’s remit and serve as a forum to share knowledge and develop software.
Proposals should be based on a clear and demonstrable need for HEC in the research area.
Each consortium will help bring HEC to a wider community, including non-traditional users and traditional experimentalists, to facilitate world-class research using the UK’s national computing resources.
Through this funding opportunity, EPSRC will aim to fund a portfolio of HEC consortia that aligns with the current digital research infrastructure landscape and continues to support and develop thriving HPC user communities.
The aims of this funding opportunity are to:
- provide key engineering and physical sciences research communities with flexible science-led access to the UK national HPC service that will enable world-class and world-leading scientific output
- support community networks that promote the highest quality science and provide leadership in developing a strategic agenda for the representative communities
- ensure greater scientific productivity on the UK national HPC service by making sure that the code base for computational science and engineering is fit for purpose, internationally competitive, responsive to technological developments and are sustainable for the future
- broaden the HPC user base and provide appropriate training, expanding the UK computational science and engineering skills base
- maximise the impact that the UK national HPC service is having by promoting strong links with industrial collaborators and users.
The consortia funded from this funding opportunity will need to create and maintain close relationships to existing activities in synergistic areas of computational research.
Proposals should, therefore, discuss the broader needs and existing funding for software development, support and maintenance. They should show how requested resources for the consortium will complement existing activities and be used effectively to build those close relationships.
Proposals submitted to this funding opportunity will need to clearly articulate why the consortium model is an appropriate one to support the users and the research community that it represents.
This will require the community, whether established or emergent, to have an appropriate level of maturity to be able to take full advantage of the resources provided as part of this funding opportunity.
In particular, proposals will need to justify:
- the benefits that the community anticipates receiving from the requested resources
- that the requested resources will be effectively utilised by the user base
- that the resources will be allocated in an inclusive manner.
Category of consortia
There will be two distinct streams for support through this funding opportunity. It will be up to you to contact EPSRC if you are unsure which category you should be applying to. The categories are as follows.
New and emerging communities consortia
This category is aimed towards communities that are already using HPC to enable research that would not otherwise be possible, even in only a limited capacity.
New and emerging communities may be small and growing or rely on computational resources and expertise that are not broadly accessible beyond individual grant awards or within a limited range of groups and institutions.
These communities will be able to:
- demonstrate a growing demand for computational resources
- evidence that the consortium will derive substantial benefit from gaining a reliable source of HPC access and technical support, which will have a transformational impact on the research capability of the community.
Established communities consortia
This category is aimed towards communities where extensive and widespread usage of the UK’s top tier HPC services is already evident. Also where usage of HPC is a well-established and crucial enabler of research.
Established communities may not be covered by an existing HEC consortia but must be able to demonstrate a critical mass of activity that justifies the substantial resources available through this funding opportunity. This includes evidence of existing collaborative activity and demand for resources.
These two categories will be assessed distinctly with the funding and HPC and technical resource split to be decided once the outline stage has been assessed.
Objectives and outcomes
The objectives and outcomes of the funded HEC consortia can be split up into the following headings:
- To serve as a forum to communicate research and HEC expertise within the relevant community. To also help UK growth in this area of research by synthesising and disseminating good practice from the global user community.
- To offer to consortia members:
- networking, mentoring and awareness of training and career development opportunities
- travel grants for code development and engagement with industries
- support towards tier 0, 1 and 2 applications
- access and core support to a suite of complementary flow solvers.
- To develop, port, optimise and benchmark software that will enable new research areas to be tackled and will prepare the UK’s user communities to benefit from future high end computing architectures.
- To expand the breadth of the work of the community that the consortium would support. Focusing on cutting-edge applications, and building collaborations with experiments and industry, to achieve maximum impact from EPSRC’s investment in the ARCHER2 and tier-2 services.
- Develop tools to assist large-scale simulation. For example, for tools to allow interoperability between different software packages, and the exchange of data and metadata and contribute to the development of data standards.
- Develop tools to facilitate analysis of the large volumes of data produced by simulations at large length and timescales.
- To provide a central core of support needed to accommodate a critical mass of activity within the community considering the big challenges posed by and to the community or
- To create and maintain information on best practice for high end computing modelling and simulation, including code-specific guidance for the community.
- To disseminate scripts, code and results within the consortium and to develop and provide advice on optimum job parameters for software on ARCHER2 and tier-2 services.
Management and governance
- To have a clear management and governance structure for the allocation and application procedures of the consortium. This must include transparent policies for allocation panel membership and open application routes for members and non-members of the consortium. Decisions should follow a peer-reviewed process that embodies EPSRC principles and standards.
- To have a flexible plan for resource utilisation over the consortium’s lifespan, aimed at maximising the consortium’s ARCHER2 and tier-2 allocation, including any changes to that allocation.
Subject to quality, funding of up to £3 million is available through this funding opportunity.
Additionally, a block of computing allocations on the UK national HPC services ARCHER2 and the portfolio of tier-2 services will be available.
Finally, technical computational support is also available through this funding opportunity.
This can either be internally sourced and costed appropriately, or available from staff at Computational Science Centre for Research (CoSeC), where up to £700,000 per annum across all HEC consortia is available.
Please make contact with CoSeC by 13 May.
We encourage applicants to continue to work with CoSeC throughout both stages of proposals.
EPSRC expects to support approximately seven to 10 developed consortia and a number of new and emerging community consortia through this funding opportunity.
HEC consortia supported through this funding opportunity must support research and research communities within the EPSRC’s remit and will have a duration of no longer than four years.
As these are community-based consortia, only one HEC consortium will be supported per thematic area.
Resources designated for the above objectives and outcomes should be requested as one of the three types of support available through this funding opportunity:
- management, administrative and networking costs
- compute allocation on the UK national HPC services
- software development and code support.
Management, administrative and networking costs
Requests for funding may include:
- investigator salaries: principal investigators and HEC consortium chairs can request funds to cover their salary costs for the time spent setting up, managing and coordinating the HEC. Requests can be made for co-investigator time, however this should be restricted to individuals that are heavily involved in the management and co-ordination of the HEC consortium and the compute allocation. Salary costs of HEC consortia members and users should not be included in the proposal
- administrative support: funding for administrative support can be requested to help in the co-ordination of the consortium and its allocation management and networking activities
- workshop and meeting costs: including funding for workshops and events, cross-consortia or collaborative computational project activities and community training
- travel and subsistence
- dissemination and communication activities: including web-based dissemination and impact activities such as collecting and publishing case studies
- industrial outreach: including short-term pilot projects and secondments.
HPC allocation request
In the outline stage, you should provide information on the requested compute allocation for the entire duration of your grant based on the existing ARCHER2 allocation unit compute units (CU) and available tier-2 resource.
ARCHER2 compute is a finite resource and the service is currently heavily utilised. In this context, you should:
- carefully consider and justify the current and future compute requirements of your members and the research communities that you represent
- specify why resource on the UK national HPC service is required.
You should also carefully consider how your consortium will manage the six monthly allocations, both in terms of allocating resource to support high quality research and managing the usage of your members.
All potential consortia must obtain a service-specific technical assessment for the requested compute resource.
EPSRC reserves the right to alter and adjust the requested compute allocations based on the available resource. This includes moving allocations between their portfolio of HPC services and delaying the start dates of projects where circumstances require.
EPSRC encourages applicants to consider whether the level of resource they are requesting is realistic and reasonable, and to refine this whilst completing their outline application.
The services available other than ARCHER2, are the EPSRC funded tier-2 HPC services:
Please see the websites of the services for further information on the capabilities available.
The indicative computational allocation available for the cohort of successful consortia is approximately 65% of EPSRC’s overall HPC resource.
This equates to around 22.5 million CU per year, measured in ARCHER2 compute units, although small amounts of graphics processing unit access are also available. Precise levels are subject to variations in the current usage of the services.
Please also note that services will expect users to either use existing software on the system or build their own software, but will provide support where possible.
If you have any queries, please contact the relevant services or view their websites when preparing your proposals, to determine what is feasible.
The start date of each funded project must be within two months from the notification of outcome.
If the relevant service is not operational at the time of notification, then applicants are expected to start their project within two months of the start of service as determined by EPSRC.
Equipment over £10,000 in value (including VAT) is not available through this funding opportunity. Smaller items of equipment (individually under £10,000) should be in the ‘directly incurred, other costs’ heading.
Learn more about EPSRC’s approach to equipment funding.
You are expected to work within the EPSRC framework for responsible innovation.
Applicants planning to include international collaborators on their proposal should visit Trusted Research for guidance on getting the most out of international collaboration whilst protecting intellectual property, sensitive research and personal information.