skip to content

IT Help and Support

University Information Services

Research Computing Services (RCS) is a centre for research, development and teaching in advanced computing systems. We design and maintain some of the world’s fastest compute and storage systems for the benefit of our University’s students and researchers, as well as industry and the wider research community. RCS is also home to a team of Research Software Engineers whose focus is on collaborating with researchers across the University to improve the quality and sustainability of research software.

RCS can facilitate your research in many ways.

High performance computing

We offer a range of advanced platforms for computationally intensive research. These are available to you in addition to your departmental resources.

Our two main high performance computing (HPC) platforms are Peta4 (a CPU cluster) and Wilkes3 (a GPU cluster). These are collectively known as the Cambridge Service for Data-Driven Discovery (CSD3). You can apply for a user account on CSD3.

Individuals and research groups at the University of Cambridge (and affiliated institutions) are eligible for an allocation of free-at-point-of-use hours on the system in each University financial quarter.

Paid access to CSD3 is also available to any group – whether from the University of Cambridge, another academic institution or from within industry – under a simple pay-for-use mechanism. These pay-for-use service tiers allow for larger, longer jobs with higher priority and a quicker turnaround time. Further information is available on the Research Computing website.

If you are an internal University of Cambridge user, you may purchase credits (in addition to the free hours you are entitled to) at any time by departmental purchase order. Further details are available on the frequently asked questions page on Research Computing Service website (Raven protected) .

If University of Cambridge users wish to include the costs of CSD3 resources on a research grant, please contact to discuss the amount of CSD3 resources you will require.

Note that EPSRC-funded University of Cambridge researchers can also apply for CSD3 resources through calls for Tier2 resources. Similarly, it may also be possible for STFC-funded researchers to apply for resources through DiRAC or IRIS National facilities for which CSD3 is a service provider. We recommend that you apply for access as a University of Cambridge user before exploring these options. You will then have convenient access to both Cambridge and National facility resources from the one user account. For more information, please contact

Secure research computing

The Secure Research Computing Platform (SRCP) is ISO 27001 certified. It provides clients with a secure, private-tenanted cloud platform for the storage and processing of level three confidential data as defined by the University-approved information security classification.

SRCP offers individuals and research groups from the University of Cambridge and other affiliated institutions access to high performance computation and data storage services within a trusted research environment.

For more information, please see our SRCP user guide or contact

Data storage

Our storage services are designed to meet the growing need for mass data storage in computationally intensive research. These services offer a competitively priced alternative to storage on local platforms. They deliver unique capabilities, including a trusted research environment, available on both the Cambridge University Data Network (CUDN) and the HPC high speed network.

Our services are designed to offer flexibility. The self-service model allows researchers, research groups and institutions to take control of the purchased data storage and allocate it as needed across projects. The self-service portal allows the purchaser to manage the storage allocations and control access to data shares.

We offer three kinds of storage to suit the various needs of our users:

  • Research Data Store (RDS) provides large-scale, high performance storage for research data. This service is ideal for individuals or research groups wishing to process data on the HPC service, or to store an additional copy of data processed elsewhere. To keep the cost to users low, RDS does not provide backups and is therefore not suitable for storing data that cannot be recovered from elsewhere or that cannot be reproduced. RDS is remotely accessible using Secure File Transfer Protocol (SFTP).
  • Research File Share (RFS) provides large-scale, enterprise-grade storage with additional data safety features. It is suitable for group shares that need additional protection and accessibility. This service delivers a highly resilient storage solution with snapshots and back-ups to allow for the fast recovery of lost files.
  • Research Cold Store (RCS) provides large-scale, long-term storage for inactive data at exceptionally low cost. It is designed to provide secure, durable storage for data packaged in larger files that can tolerate longer retrieval times. It is accessible remotely using the Secure File Transfer Protocol (SFTP).

We offer dedicated technical support through the our service desk at

Visit our site to learn more about our research data storage services.

Research software engineering

RCS is home to a team of research software engineers who combine expertise in programming with an intricate understanding of the university research environment. Our Research Software Engineering (RSE) team works in close collaboration with researchers, research groups and industry to design and produce high quality software for research. They deliver training and workshops on programming for HPC, as well as providing advice and consultancy.

To discuss a project or to request further information, please contact


We can provide training in high performance computing and high quality software engineering. To discuss your training needs or to request further information, please contact


We can provide free consultancy support. We have a wealth of expertise within the department and would be glad to answer questions and provide one-to-one meetings with researchers who want to make use of HPC systems, develop a piece of research software or improve the quality of software already engineered. If you would like to arrange a consultancy session, please contact