Hardware
========

The *entropy* cluster consists of 12 compute nodes.

.. _hardware_anchor:

Compute nodes
-------------

.. csv-table::
   :header: "Server", "CPU", "GPU", "Memory"

   "a100a", "AMD EPYC 7713", "8x A100", "1024 GB"
   "h100a", "AMD EPYC 9534", "8x H100 HBM3", "1024 GB"
   "asusgpu1", "Intel Xeon E5-2640 v4", "8x Titan V", "128 GB"
   "asusgpu2", "Intel Xeon E5-2660 v4", "8x RTX 2080 Ti", "128 GB"
   "asusgpu3", "Intel Xeon E5-2660 v4", "2x RTX 1080 Ti, 6x Titan V", 128 GB
   "asusgpu4", "Intel Xeon E5-2660 v4", "2x RTX 2080 Ti, 6x Titan X", 128 GB
   "asusgpu5", "Intel Xeon E5-2640 v4", "2x RTX 2080 Ti, 6x Titan X", 128 GB
   "asusgpu6", "Intel Xeon E5-2640 v4", "8x RTX 2080 Ti", "128 GB"
   "arnold", "Intel Xeon E5-2630 v4", "4x Titan V", "64 GB"
   "bruce", "Intel Xeon E5-2630 v4", "2x RTX A6000", "128 GB"
   "steven", "Intel Xeon E5-2630 v4", "4x Titan V", "64 GB"
   "sylvester", "Intel Xeon E5-2630 v4", "4x Titan V", "64 GB"

GPUs
----

The cluster has several models of nVidia GPUs. They can be selected for jobs
using ``--gres=gpu:gres_type:amount`` (please refer to the :ref:`submitting_jobs`).

.. csv-table::
   :header: "GPU model", "GRES type"

   "GTX 1080 Ti", "1080ti"
   "Titan X", "titanx"
   "Titan V", "titanv"
   "RTX 2080 Ti", "rtx2080ti"
   "RTX A6000", "a6000"
   "A100", "a100"
   "H100", "h100"