Computer Clusters

On this page

Physical Clusters

ITS provides one site where ECE students may sit at the computing device. This is in Hamerschlag Hall 1305 and consists of 20 Dell Optiplex 760 workstations. Each unit is equipped with a 24″ monitor.

The machines found in HH1305 are:

[apollo, ares, artemis, cronus, demeter, erida, erinyes, eris, fates, gorgons, graces, hades, hebe, helios, hermes, hestia, leto, nemesis, oceanus, themis]

Remote Access Clusters

There are several remote access compute clusters located in the ITS Cyert Hall machine room, some of which are publicly accessible and some of which are only accessible from within the CMU network. They can be accessed remotely either directly through ssh, sftp, or, indirectly, via a service such as condor. Condor is described in greater detail elsewhere in this guide.

    • General purpose connection server – not for computationally intensive work
    • Hostnames
    • The primary private site is known as the ECE compute cluster with 32 machines. All faculty, grad & undergrad students automatically receive access to these machines. Machine names ece000 through ece031.


Hostnames Cores RAM (GB)
ece000 thru 24 64
ece008 thru ece013.ece.local.cmu.ed 20 128
ece014 thru 28 256
ece017 thru 16 16
ece023 thru 20 128
ece029 thru 28 256
    • A small number of graphical Linux servers are available via CMU networks only using FastX


    • Various research groups or principle investigators have their own compute clusters. Access to those clusters is only available with the consent of the party who owns the cluster.
    • ECE Community members interested in using the ECE Data Science Cloud (DSC) must apply for access using this form.
    • Only ECE graduate students and faculty may use DSC systems. Approval is granted based on available resources and the type of research being conducted. Students must provide an email contact for their advisor so ECE IT Services may obtain their consent.
    • The DSC is comprised of systems with larger memory, gpu, and cpu resources than standard servers. Large simulation jobs must be submitted via a Condor head node.
    • Please submit inquiries using the ECE Data Science Cloud Inquiry form