Linux Compute Servers

List of Servers

Name Alias Hardware Operating System
linuxg4 linux18 1 x AMD EPYC 7662 64-Core Processor 128GB RAM
1 x NVIDIA Tesla A100 GPU
CentOS 7 GPU set in multi-instance mode. See here for more info.
linuxg3 linux17 1 x AMD EPYC 16-core with 64GB RAM
2 x NVIDIA Tesla V100S GPU
Rocky Linux 8
linuxg2 linux16 1 x Intel Xeon E5-2620 v4 8-core (2.1GHz) with 64GB RAM
2 x NVIDIA Tesla P100 GPU
CentOS 7
linuxg1 linux15 1 x Intel Xeon E5-2683 16-core (2.1GHz) with 128GB RAM
1 x NVIDIA Tesla P100 GPU
CentOS 7
linuxf4 linux14 2 x Intel Xeon E5-2603 Six core (1.7GHz) with 128GB RAM CentOS 7
linuxf3 linux13 2 x Intel Xeon E5-2603 Six core (1.7GHz) with 128GB RAM CentOS 7
linuxf2 linux12 2 x Intel Xeon E5-2603 Six core (1.7GHz) with 128GB RAM CentOS 7
linuxf1 linux11 2 x Intel Xeon E5-2603 Six core (1.7GHz) with 128GB RAM CentOS 7
linuxe4 linux10 2 x Intel Xeon E5-2630 Six core (2.6GHz) with 64GB RAM CentOS 8
linuxe3 linux9 2 x Intel Xeon E5-2630 Six core (2.6GHz) with 64GB RAM Rocky Linux 8
linuxe2 linux8 2 x Intel Xeon E5-2630 Six core (2.6GHz) with 64GB RAM Rocky Linux 8
linuxe1 linux7 2 x Intel Xeon E5-2630 Six core (2.6GHz) with 64GB RAM Rocky Linux 8
linuxb1 linux6 1 x AMD EPYC 7351 16 Core (2.4GHz) with 128GB RAM CentOS 7
linuxa5 linux5 2 x Intel Xeon Fourteen Core (2.4GHz) with 256GB RAM CentOS 7
linuxa4 linux4 2 x Intel Xeon Fourteen Core (2.4GHz) with 256GB RAM CentOS 7
linuxa3 linux3 2 x Intel Xeon Fourteen Core (2.4GHz) with 256GB RAM CentOS 7
linuxa2 linux2 2 x Intel Xeon Fourteen Core (2.4GHz) with 256GB RAM CentOS 7
linuxa1 linux1 2 x Intel Xeon Fourteen Core (2.4GHz) with 256GB RAM CentOS 7

Software

Server programs are installed at the request of Faculty and Staff. Some general software is available such as:

  • ANSYS
  • MATLAB
  • Common GNU compilers such as c/c++ and g77.
  • Anaconda and Python 2.7 and 3.X

Most of the software can be found in the usual directory locations and /usr/local, but contact us for more info or for more information on specific program availability if needed.

Multi-Instance GPU

The Multi-Instance GPU (MIG) feature allows GPUs be partitioned into separate instances. To identify and choose the instance you’d like to use:


1. Identify the instances available:

    nvidia-smi -L

2. Set your CUDA_VISIBLE_DEVICES environment variable to the instance you’d like to run on.

    If using tcsh

    setenv CUDA_VISIBLE_DEVICES MIG-xxx-xxx-xxx-xxx-xxx

    If using bash

    export CUDA_VISIBLE_DEVICES=MIG-xxx-xxx-xxx-xxx-xxx

Then run your code as normal. For more information, see here.

Walter Scott, Jr. College of Engineering logo

Engineering Technology Services

Bookmark and use our online help desk form first.
 

Email help@engr.colostate.edu
Call (970) 491-2917
 

Stop by an ETS Help Desk:
Main Help Desk (Glover 100)
Foothills Campus Help Desk (Atmospheric Science 107)
Academic Village Help Desk (AVB C142, next to the Orion Design Studios)