> home > about > events > using_hpcx
Using the HPCx Service

The Workshop

HPCx is the UK's newest and largest National High Performance Computing system, comprising 1280 IBM POWER4 processors and delivering up to 3.4 TeraFlops/s sustained performance. It is currently ranked number 11 in the top 500 supercomputers (See: www.top500.org)). This system has been funded by the the Engineering and Physical Sciences Research Council (EPSRC). The project is run by the HPCx Consortium, a collaboration comprising the University of Edinburgh, Edinburgh Parallel Computing Centre (EPCC), CCLRC's Daresbury Laboratory and IBM.

This aim of this workshop is to provide an overview of the HPCx service and the current challenges associated with the system. The sessions will consist of three main topics: an overview of the system; terascaling of real applications; and consideration of the performance issues surrounding the use of HPCx.

Presentations

The HPCx Service

This talk will provide an introduction to HPCx, the UK's new National HPC Service. The aim of the service is to deliver a world-class service for capability computing to the UK scientific community, and part of the talk will focus on the challenges associated with achieving such a goal. The remainder of the talk will give an overview of the architecture, highlighting some of the issues relevant to users.

Achieving terascaling performance from capability applications

This talk will focus on the CRYSTAL code. Ian will discuss the capabilities of the MPP CRYSTAL application and summarise the algorithms used within the code and how they were parallelised. The scaling of the code will be illustrated. These results will show that increasing the problem size increases the scalability of the code, and that for the largest system very good scaling up to 1024 processors is obtained ( ~700 speed-up on 1024 procs ).

Capability Computing: Achieving scalability on HPCx

The main objective of HPCx is to provide a capability computing service for a range of key scientific applications, i.e. a service for applications that can utilise a significant fraction of the resource. To achieve this capability computing objective, applications must be able to scale effectively to around 1000 processors. This presents a considerable challenge, and requires an understanding of the system and its bottlenecks. In this talk Gavin will present results from a detailed performance investigation on HPCx, highlighting potential bottlenecks for applications and how these may be avoided.

Registration