The High Performance Computing (HPC) service has three supercomputers, housed in Imperial College London's Data Centre. These systems provide substantial computing capacity to researchers in College.
What we offer
- Cx1 (general purpose computer cluster) covers a significant amount of the computing requirement with a large and varied workload ranging from myriad small runs, such as parameter studies, to long-running CPU intensive runs and small-scale parallel runs, achieving a high output of calculations. The architecture of the system is aimed to provide as much computing power at the lowest possible cost, with new technology added as it becomes available to keep hardware at the frontline of processing capacity.
- Cx2 (high end parallel system) is a large scale, massive parallel processing (MPP) system. It was specifically acquired to run parallel programs using hundreds or thousands of cores on a single run and is reserved for users whose programs scale to the large number of central processing units (CPUs).
- Ax3 (large shared memory system) is a very large shared with restricted access for those requiring specialised features - currently for Genomics and Chemistry research.
Data Centre hosting
Our HPC service offers option for hosting HPC equipment purchased by departments in the Data Centre, including:
- The housing of systems and the provision of power and cooling.
- Assistance with planning, configuration and installation.
- Help with software configuration.
- System monitoring.
- System management and operational support.
- Fully managed system hosting including software and operational support.
Options are designed to take advantage of the economies of scales and improve efficiency, reducing the hosting burden and cost of HPC system to departments.
Who runs the service?
Our HPC Service Support team runs, maintains and administers HPC facilities.