Difference between revisions of "CAD and Visualization Laboratory"

From Gobblerpedia
Jump to: navigation, search
imported>Matthazinski
(Added details about giant printer)
imported>Matthazinski
(Compute Clusters: Corrected number of cluster nodes)
Line 11: Line 11:
  
 
== Compute Clusters ==
 
== Compute Clusters ==
* 16-node old cluster
+
* <strike>15</strike> 14-node old cluster
 
** Each has Intel(R) Pentium(R) D CPU 3.00GHz CPUs, 8 GBytes RAM, 1 Gbit/sec ethernet  
 
** Each has Intel(R) Pentium(R) D CPU 3.00GHz CPUs, 8 GBytes RAM, 1 Gbit/sec ethernet  
 
** Remote access available through ssh to <code>cvl.ece.vt.edu</code>. Individual nodes are <code>cvlXX.ece.vt.edu</code>
 
** Remote access available through ssh to <code>cvl.ece.vt.edu</code>. Individual nodes are <code>cvlXX.ece.vt.edu</code>
 
** MPI is installed, qsub can be used to parallelize things
 
** MPI is installed, qsub can be used to parallelize things
 
** X11 is also installed, so <code>ssh -X cvl.ece.vt.edu</code> can be used to remotely run GUI applications.
 
** X11 is also installed, so <code>ssh -X cvl.ece.vt.edu</code> can be used to remotely run GUI applications.
* 32-node blade cluster
+
* <strike>32</strike> 31-node blade cluster
 
** 16 nodes with 2x Quad-Core AMD Opteron(tm) Processor 2350 2GHz, 20 GBytes RAM, 1 Gbit/sec ethernet (<code>cluster[01-16]</code>)
 
** 16 nodes with 2x Quad-Core AMD Opteron(tm) Processor 2350 2GHz, 20 GBytes RAM, 1 Gbit/sec ethernet (<code>cluster[01-16]</code>)
 
** 16 nodes with 2x Quad-Core (8 threads) Intel(R) Xeon Processor E5620 2.40 GHz, 24 GBytes RAM, 1Gbit/sec ethernet (<code>cluster[17-32]</code>)
 
** 16 nodes with 2x Quad-Core (8 threads) Intel(R) Xeon Processor E5620 2.40 GHz, 24 GBytes RAM, 1Gbit/sec ethernet (<code>cluster[17-32]</code>)

Revision as of 18:22, 28 January 2013

The CAD and Visualization Laboratory or CVL is a computer lab running mainly CentOS machines maintained by the Bradley Department of Electrical and Computer Engineering. It is located in 432 Whittemore Hall. Access is through a keypad, the number to which can be obtained through official channels from ECE systems administrator Branden McKagen. Disk quotas are enforced when home folders become unreasonably large, unless you have a good reason to need the space. The CVL cluster is available via ssh to account holders on cvl.ece.vt.edu and has MPI, MATLAB, Cadence, and other software installed.

Virginia Tech users can obtain a central authentication account for all CVL machines using the automated account creation tool.

Physical Resources in the CVL

The lab in 432 Whittemore Hall contains:

  • 9 workstations (cvlwsXX.ece.vt.edu), most of which have CentOS, Core 2 Duo, 4 GB memory, and dual monitors
  • A printer available at socket://172.16.12.1 (only accessible from machines behind the CVL router). This uses the Dell M5210 Foomatic/Postscript driver, which is available in /software/Dell_5210n.ppd on CVL machines.
  • A microscope
  • A large-format Epson 9800 printer for research groups to print posters and CAD drawings (located offsite, follow the posted instructions for usage).

Compute Clusters

  • 15 14-node old cluster
    • Each has Intel(R) Pentium(R) D CPU 3.00GHz CPUs, 8 GBytes RAM, 1 Gbit/sec ethernet
    • Remote access available through ssh to cvl.ece.vt.edu. Individual nodes are cvlXX.ece.vt.edu
    • MPI is installed, qsub can be used to parallelize things
    • X11 is also installed, so ssh -X cvl.ece.vt.edu can be used to remotely run GUI applications.
  • 32 31-node blade cluster
    • 16 nodes with 2x Quad-Core AMD Opteron(tm) Processor 2350 2GHz, 20 GBytes RAM, 1 Gbit/sec ethernet (cluster[01-16])
    • 16 nodes with 2x Quad-Core (8 threads) Intel(R) Xeon Processor E5620 2.40 GHz, 24 GBytes RAM, 1Gbit/sec ethernet (cluster[17-32])
    • This cluster does not have global IPv4 addresses, so you must ssh into other machines first to access this.
  • Both clusters use a shared file storage system and central authentication, so home directories will persist between nodes.
  • Lots of software is installed on the CVL machines.

For larger jobs, the Advanced Research Computing machines may be more appropriate.

External Links