Report on Computational Science and Engineering Support Activities at Daresbury Laboratory 1999/2000 icon

Report on Computational Science and Engineering Support Activities at Daresbury Laboratory 1999/2000


1 чел. помогло.
Similar
Report on Computational Science and Engineering Support Activities at Daresbury Laboratory...
Center for Vision Science and Technology and Visual Processing Laboratory...
Computational Neuroethology Laboratory...
Naval Research Laboratory, Center for Bio/Molecular Science and Engineering, Rm. 406 Bldg. 30...
As Chairman, I now present our annual report on the activities of the charity...
B. Sc. Degree in Electrical Engineering, University of Engineering and Technology Lahore...
Japan Science and Technology Agency icorp, Computational Brain Project...
Using Compendium as a tool to support the design of learning activities...
Professor of Chemical Engineering and Director of Polymer Science and Engineering...
Template for proposing new mars science laboratory landing sites...
Report of fep activities & European Events...
Proceedings of the International Conference...



страницы: 1   2   3   4   5   6   7   8   9   ...   70
return to the beginning

Introduction





This report describes work done for EPSRC under the Facilities Agreement 'Computational Science and Engineering Support and Development Activities at Daresbury Laboratory 1999/2000'. Its structure follows the headings in the SLA, namely:




  • Hardware, Operations and Computing Infrastructure;

  • CCP Support;

  • HPC Development;

  • Distributed Computing; and,

  • Finance.




The report addresses the various activities detailed in the SLA with a brief report on progress and pointers to fuller technical accounts of the work in the various Annexes. These technical reports will appear as part of the Computational Science and Engineering Department's Annual Report for 2000/2001 in due course.


Over the past year the major activities have been:




  • Upgrading the IBM/SP2 system with 32 processor Winterhawk-2 processors.

  • Installation of a 32 processor Intel Pentium Beowulf-class system.

  • Significant progress in reviewing the strategy for the Collaborative Computational Projects (CCPs) due for completion in the summer of 2000.

  • Preparing science and engineering case studies for the HPCx business plan.

  • Major advances in the development of new functionality for and restructuring of for the core academic codes such as CRYSTAL, CASTEP, GAMESS-UK and DL_POLY.

  • Extensive performance evaluation of ASCI class and Beowulf systems using real applications.




A quarterly newsletter, HPCProfile, is produced by the Department which includes information of relevance to all of the HPC and computational science activities in the UK including distributed computing, CCPs, High-End Computing and HPCI. Editions over the last year include:




  • June 1999: featuring articles on the parallelisation of an external aerodynamics code FLITE3D, parallel programming with pthreads and an overview of the distributed computing support programme;

  • September 1999: a special issue on numerical algorithm research for high-performance computing; and,

  • December 1999: which featured articles on tuning DL_POLY for vector computers and an ab initio estimate of the temperature of the Earth's core.




As of January 2000, the HPC Profile will concentrate on disseminating information about the CCP and DisCo programmes through a biannual publication. The High End Computing Collaboration will disseminate information about its activities through a separate newsletter.


^

Hardware, Operations and Computing Infrastructure




Hardware Resources





Major hardware resources currently at Daresbury include:


IBM RS6000-SP: 26-processor system with peak speed of 12.0536 Gflops, aggregate memory of 6.4 GB, aggregate disk capacity of 118 GB. The detailed configuration is:

  • 2 SP2 Wide Nodes (66.7 MHz) each with 128 MB memory;

  • 24 SP2-SC Thin Nodes (120 MHz) each with 256 MB memory;

  • 3 GB disk capacity local to each compute node; and,

  • 40 GB disk capacity on an attached RAID3 system.




Beowulf Pilot System: 8-node system based on Pentium processors interconnected via standard ethernet.


Miscellaneous desktop systems: PCs and IBM, SGI, DEC-AXP, and SUN desktop workstations are used for code development, testing etc.


During the year EPSRC approved funding for a proposal to enhance the IBM SP2 with the provision of ASCI class nodes and the provision of a 32-processor Beowulf system. Hence the hardware base noted above was expanded to include:


IBM -SP 32 processor Winterhawk-2 System: with the following configuration:

  • 8 Winterhawk-2 nodes, each equipped with four Winterhawk-2 processors, 2GB of memory and 18.2 GB of disk;

  • 147 of SSA disk comprising 8 18.2 GB disks;

  • 2 silver nodes (2 332 MHz 604e processors, 512MB of memory, 4.5GB +9.1 GB disks), one acting as a GFPS server and the other enhancing the login nodes; and,

  • a new 9.1GB disk for control workstation.



32 processor Intel Pentium system:

  • Master node: Intel PentiumII-266 CPU, 256MB SDRAM, 2 PCI Fast Ethernet interconnects, 4GB system and 9GB user ultrawide SCSI disks;

  • 32 processor nodes: each with Intel PentiumIII-450 CPU, 256 MB SDRAM, 2 PCI FastEthernet interconnects, 10GB U-DMA disk; and,

  • 2 extreme summit48 Fast Ethernet switches.



^

Operational Support for the IBM RS6000-SP System





2 work years of staff are used in operational support for the IBM system. Recurrent expenditure includes system maintenance and associated running costs- for newsletters, documentation etc., and for miscellaneous hardware and software purchases.


Significant activities over the next year will be:




  • Upgrade of IBM SP2 software and Y2K testing of current hardware configuration (June ‘99).

  • Installation of IBM upgrade (September ‘99), subsequent commissioning and Y2K compliance testing.

  • User comments on service levels (March ’00).



The principal task over the last year has been to maintain the systems as a reliable development and evaluation platform capable of underpinning almost all of the work covered within this and subsequent agreements. Over the past year CLRC has restructured the user base around the SP2 to reflect its prime role as a development platform. Its use as a partial production resource to the external user community has been discontinued. The user base on the enhanced IBM SP2 system includes the staff supported through this and subsequent Agreements and external users collaborating on these projects. Allocations to the various projects for the past year, and usage are listed in the table in Annex 1. Industrial usage on the SP2 over the period of the agreement has been nil.


The service levels around the IBM system have been structured to meet the requirements of a development environment and the demands of a technically sophisticated user-base. When the upgrade to the facility was introduced the rapid dissemination of information was important. The Web was used for this purpose, carefully structuring and managing the information provided. The Internet resource also acts as a depository for user documentation, programming examples, FAQs and general announcements. In the longer term the Web interface can be developed to provide interactive access to system status, job submission etc. Annex 1 presents a report on the service levels around the SP2 and how they have been met over the past year. It is worth noting that upgrades of the software ensured a smooth transition into the new millennium. The system was upgraded towards end of January with the 8 Winterhawk-2 nodes. Problems with the processor card interfaces were experienced, these were subsequently replaced and the disk system upgraded to meet the tender specification.


SP2 users met in December 1999 to discuss the service and software requirements for the enhanced system. The Forum prioritised software requirements and discussed the operational aspects required for a flexible execution environment covering both interactive and batch access. The next meeting of the Forum will take place after the enhanced system has been accepted. The service levels and operational requirements will be reviewed with user input on a periodic basis.


CLRC will make available training materials for users seeking to optimise their applications on the enhanced service. These materials will de developed using effort drawn from the HPC Development activity. See the section on ‘Applications on ASCI-class Systems’ for details of deliverables.







Download 2.54 Mb.
leave a comment
Page2/70
Date conversion31.08.2011
Size2.54 Mb.
TypeДокументы, Educational materials
Add document to your blog or website

страницы: 1   2   3   4   5   6   7   8   9   ...   70
плохо
  1
Your rate:
Place this button on your site:
docs.exdat.com

The database is protected by copyright ©exdat 2000-2017
При копировании материала укажите ссылку
send message
Documents

upload
Documents

Рейтинг@Mail.ru
наверх