Images of climate innovation
ARCHER2: Facilitating world-class climate research.
ARCHER2 is the UK's National Supercomputer, a world-class advanced computing resource for researchers across the UK. Funded by UK Research and Innovation (UKRI) and operated by EPCC at the University of Edinburgh, ARCHER2 is used by UK researchers working across a range of climate research areas including: modelling and predicting climate change; ocean research; atmospheric research, including air quality; modelling clean energy technologies such as ocean and wind.
The ARCHER2 supercomputer is an HPE Cray EX supercomputer. Capable at its peak of over eleven times the science throughput of ARCHER, the previous National Service, ARCHER2 will provide a step-change in performance for climate researchers, allowing them for example to run larger, more complex simulations than ever before.
This in turn will enhance our understanding of climate change and future energy needs, ultimately helping society adjust and adapt to a changing world. Recent examples of research enabled by ARCHER2 include improving cloud modelling for better weather and climate prediction and forecasting the dispersion of volcanic ash and gas. ARCHER2 is housed at EPCC's Advanced Computing Facility (ACF), which provides state-of-the-art facilities for large-scale systems.
Supercomputers generate significant amounts of heat, so it is essential to minimise the energy used to cool them down. The ACF has been designed to cool these systems in the most energy-efficient way possible, by pumping water to the roof of the building to allow cold Scottish air to cool it, and it is only on extremely warm summer days that there is no free cooling.
As water is a significantly more effective coolant than air alone, the ACF is a highly efficient data centre. ARCHER2 is hosted and managed by EPCC, the supercomputing centre at the University of Edinburgh, on behalf of UKRI. EPCC also supports the users of ARCHER2, ensuring they can maximise the science achieved on the service.
Entrant: Tracy Peet , University of Edinburgh
Copyright: UKRI Engineering and Physical Sciences Research Council