High Performance Computing

HKHPC supports High Performance Computing (HPC) service for University researchers and scholars from diverse disciplines whose work requires intensive computations. HPC service is currently supported by two high performance Linux clusters. These clusters are used for solving complex computational problems that cannot be supported by standard desktop computers or workstations. The clusters can support research projects that require a sizable CPU power, memory, network and disk space as well as parallel programming for improving research efficiency.  

HKHPC arranges training courses on parallel programming for staff and students to acquaint them on how to use parallel programming to optimize the use of computational power in calculations and program running.

Users are requested to acknowledge the use of HKHPC HPC facilities in their research publications. Please see Acknowledgement for more information.

HKHPC provide high-performance computing (HPC) service to University researchers.  The HPC service in University is organized base on the Purdue community cluster model, which is widely adopted by research universities all over the world.   The University research community cluster solicit Schools or Departments to contribute funding for procurement of high performance computing machines, while HKHPC would assist in coordinating joint purchase, provides the necessary infrastructure, managed data center hosting service, cluster system administration services and hardware maintenance of the machines.  Contributors do not solely own the hardware resources while the machines are in the community cluster, instead they have priority to use their contributed resources.  Whenever there are idle resources, these spare resources would be made available to other users of the HPC cluster to maximize the overall resource utilization.  The machines in the HPC cluster runs with a lifespan of five years and would be subjected to the retirement review from the 6th year onward to redistribute the scarce resources of data center space to new contributor when needed.