Purdue Homepage | Purdue Search | Campus Map | Purdue Directories
Direct inquiries or comments to: rcac-info@purdue.edu
All content Copyright © 2007, Purdue University, all rights reserved.
An equal access/equal opportunity university.
Purdue has two major TeraGrid compute resources, the Steele Cluster and the Condor Pool, as well as FPGA, file storage, data, and visualization resources.
This cluster has 893 dual quad-core Intel E5410 processor compute nodes, all running Red Hat Enterprise Linux, version 4 (RHEL4). Each of the nodes has eight 64-bit 2.33 GHz Dell 1950 CPUs, and either 16 GB or 32 GB of RAM. Some of the nodes in this cluster are interconnected with Infiniband, and most with Gigabit Ethernet. Steele users have access to a 1.3 PB DXUL archive system. Steele's peak performance is rated at 66.59 TFLOPS, and we believe it is well suited for a wide range of both serial and parallel jobs.
| SSH / Login: | tg-login.purdue.teragrid.org |
| GridFTP: | tg-steele.purdue.teragrid.org (non-striped) tg-data.purdue.teragrid.org (striped) |
| GRAM: | tg-steele.purdue.teragrid.org/jobmanager-pbs |
| Compilers: | Intel, PGI, GNU (use softenv to select) |
| PBS Queue: | tg_workq |
The Purdue Condor pool consists of over 14,000 CPUs. Of these, more than 6,500 are Linux/x86_64 CPUs, approximately 800 are Linux/Intel (ia32) CPUs, and 1,700 WinNT51/Intel CPUs. There are also small numbers of Itanium Linux, Solaris and Mac OSX machines. Memory on compute nodes ranges from 512 MB to 16 GB, and most CPUs run at 3 GHz or better. With a total of over 25 TFLOPS available, the Purdue Condor pools can provide very large numbers of cycles in a short amount of time. All shared areas and software packages available on Steele are also available on Condor. Condor is designed for high-throughput computing, and is excellent for parameter sweeps, Monte Carlo simulation, or most any serial application. Also, some classes of parallel jobs (master-worker) may be run via Condor.
| GridFTP: | tg-condor.purdue.teragrid.org (non-striped) tg-data.purdue.teragrid.org (striped) |
| GRAM: | tg-condor.purdue.teragrid.org/jobmanager-condor |
Purdue provides limited FPGA resources to TeraGrid users. These resources consist of an SGI 450 (brutus.rcac.purdue.edu) with two RC100 FPGA blades, totaling 4 available FPGAs. Also available is a Sun Fire X2200 M2 (portia.rcac.purdue.edu) which serves both as a place & route node for preparing FPGA code for use on Brutus and as an entry point for GSI-SSH and job submission to Brutus by TeraGrid users.
| SSH / Login: | fpga.purdue.teragrid.org |
| Place & Route Software: | Mitrionics + Xilinx |
Purdue provides a cloud computing testbed "Wispy" to TeraGrid users. It consists on one frontend and VM image storage nodes, and four dual-CPU VM host machines. The machines have one and a half gigabytes of available memory.
The cloud supports virtual machines with real Internet addresses, so researchers are able to get running in the cloud with minimal complications. Anyone interested in an account on our cloud should send justification and an X.509 DN to rcac-help@purdue.edu.
| Cloud Quick Start Guide (globus.org) |
| No VPN software is necessary. |
| Wispy works with version 9 of the workspace client |
| Configuration file is located here (remove .txt to use) |
| Nimbus Software Release T2.1 |
All TeraGrid users at Purdue have a home directory, which is the default directory you are placed in when you log in. This is where you should store files you want to keep over a long term such as source code, scripts, input data sets, etc. It should also be used for files you need to use often. Your home directory physically resides on a BlueArc Titan 2500 system, and this is backed up regularly. Aside from a home directory, all users have access to a scratch directory, which is pre-created for everyone with a quota of 250 GB. Note that no backup is made of this scratch space, and it is purged after 60 days. The scratch directories are also stored on a BlueArc server. User scratch directories are located in subdirectories on the scratch95 or scratch96 filesystems with names beginning with the first letter of the user's login name.
When referring to your scratch space in scripts, always use the variable $RCAC_SCRATCH. The specific path may change at any time, but this variable will always point to the correct location.