CI Survey: South Dakota State University

From GPNWiki

Jump to: navigation, search
  '''  on either side of text will bold it. Please '''bold''' your answers. 

Contents

Submitted by

Del Johnson

CI Capability (Based on NSF 2007 Survey Questions)

  • Total bandwidth to both I2 and commodity Internet
  • Commodity Internet
    • bandwidth 2007 100 mbps
    • bandwidth 2008 155 mbps
    • Number of lines 2007 1
    • Number of lines 2008 1
  • Internet 2/Abilene bandwidth
    • bandwidth 2007 155 mbps
    • bandwidth 2008 10 gbps
  • Did any of your bandwidth come from a consortium?
    • 2007 yes
    • 2008 yes
  • Do you have any connections to high performance networks?
    • 2007
      • Internet 2 yes
      • National Lamba Rail no
      • Fed Gov research network (DoE ESNet, NASA NREN)no
      • State or regional high performance network yes
      • Other, please specify no
    • 2008
      • Internet 2 yes
      • National Lamba Rail no
      • Fed Gov research network (DoE ESNet, NASA NREN) no
      • State or regional high performance network yes
      • Other, please specify no
  • What speed are your desktop ports connected to the campus?
    • 2007 100 mbps
    • 2008 100 mbps (95%) 1 gbps (5%)
  • Does your institution have any dark fiber (ie, fiber laid, but not in use)?
    • 2007 yes, between buildings only
    • 2008 yes, between buildings only
  • What was the distribution (backbone) speed of a campus computer to the campus network?
    • 2007 1 gbps
    • 2008 10 gbps
  • What percentage of campus was covered by wireless?
    • 2007 31-40%
    • 2008 61-70%
  • Did your campus provide central High Performance Computing of 1 teraflop or faster?
    • 2007 no (less than 1 TF)
    • 2008 yes
  • How many single core computing clusters (and size) did your institution have?
    • 2007 1 - .35 TF
    • 2008 1 - .35 TF
  • How many dual core computing clusters (and size) did your institution have?
    • 2007 none
    • 2008 1 - .70 TF (includes some quad core blades)
  • How many quad core computing clusters (and size) did your institution have?
    • 2007 none
    • 2008 1 - .70 TF (includes some dual core blades)
  • How many 8-core computing clusters (and size) did your institution have?
    • 2007 none
    • 2008 none
  • Please provide any other information about your campus cluster you'd like.

The SD Board of Regents is proposing a significant upgrade for cluster computing. If this is able to move forward, SDSU should be able to move to a 2-3 TF system in the next year.

  • How many SMP at 1 teraflop or faster did your institution have?
    • 2007 none at 1 TF, 2 at significantly less than 1 TF
    • 2008 none at 1 TF, 4 at less than 1 TF
  • What was the theoretical peak performance of your fastest SMP system of 1 teraflop or faster?
    • 2007 none at 1 TF
    • 2008 none at 1 TF
  • How many PVP systems of at least 1 teraflop did your institution have?
    • 2007 none
    • 2008 none
  • What was the theoretical peak performance of a)your fastest cluster, and b)all clusters of 1 teraflop or faster?
    • 2007
      • a .35 TF
          • 2008
      • a .70 TF
      • b .70 TF
  • How many MPP systems of 1 teraflop or faster did your institution administer?
    • 2007 none
    • 2008 none
  • What was the theoretical peak performance of a)your fastest MPP of 1 teraflop, and b)all MPP of 1 teraflop?
    • 2007
      • a none
      • b none
    • 2008
      • a none
      • b none
  • What was the theoretical peak performance of all PVP systems of 1 teraflop or faster?
    • 2007 none
    • 2008 none
  • Did your institution use any HPC for administrative functions?
    • 2007 no
    • 2008 no
  • How many experimental/emerging computing systems of 1 teraflop of faster did your institution administrate?
    • 2007 none
    • 2008 none
  • How many special purpose computing systems of 1 teraflop or faster did your institution administer?
    • 2007 none
    • 2008 none
  • Did any external users use your institution's high performance computing resources?
    • 2007 no
    • 2008 yes
  • What was the total available online storage for centrally administered high performance computing resources?
    • 2007 10 TB
    • 2008 10 TB
  • Out of the previous question's storage, how much was shared?
    • 2007 8 TB
    • 2008 8 TB
  • Was any of this online, shared storage used for administrative purposes?
    • 2007 no
    • 2008 no
  • What was total archival storage available specifically for centrally administrated HPC?
    • 2007 .4 PB tape library
    • 2008 .5 PB tape library
  • Was any of the previous archival storage used for university administration?
    • 2007 yes
    • 2008 yes
  • What was your total air conditioned machine room space?
    • 2007 1848 sq ft
    • 2008 1848 sq ft
  • Any additional comments you'd like to make?

==Research Programs Photovoltaics GIS Genetic studies Geonmics/Protiomics Draught Center Biofuels Sun Grant Nano Chemistry Pharmaceuticals Math/Stat Engineering Plant Science Animal and Range Science NEON

Software (e.g., Shibboleth, Condor)

Condor under development

Shibboleth Implementation

no current Shibboleth implementation

Datasets

no major datasets

Applications (e.g., BLAST for bioinformatics, WARP for weather)

SOLAR for genetic tracking, Ansys and Fluent for Mechanical Engineering, Shroedinger for Chemistry

Expertise (CI)

Networking, clusters (Linux operating system), SMP systems (AIX operating system)

Science News (RSS)