/
ECE Community Compute Clusters

ECE Community Compute Clusters

The ECE Department maintains compute resource for its graduate students and research groups.

Connection methods

Method

Description

Recommended client software

Method

Description

Recommended client software

SFTP/SCP (via SSH)

Secure File Transfer Protocol (SSH subsystem)

Used for file transfers

SSH

Secure Shell command interface

X-Session

X-Server graphical client

StarNet FastX

Advanced X-Server graphical client, allows reconnections.

fileexchange.cit.cmu.edu is for legacy ECE AFS accounts. See authentication for more information.

ECE Community Cluster Specifications (ECE NUMBER Cluster)

All fully qualified host domains are “ece.local.cmu.edu” unless specified otherwise

A connection to the CMU VPN is required.

All systems accept CMU Andrew authentication accounts.

System Group

OS

Access Methods

GPU

Cores

Memory (GB)

Scratch Storage (GB)

ece000-ece005.ece.local.cmu.edu

RHEL8

ssh

NVidia Tesla T4

12

64

512

ece006-ece013.ece.local.cmu.edu

RHEL8

ssh

NVidia Tesla T4

12

128

512

ece014-ece016.ece.local.cmu.edu

RHEL8

ssh

NVidia Tesla T4

16

256

512

ece017-ece028.ece.local.cmu.edu

RHEL8

ssh

NVidia Tesla T4

16

128

512

ece029-ece031.ece.local.cmu.edu

RHEL8

ssh

NVidia Tesla T4

16

256

512

 

 

 

 

 

 

 

ece-gui-000->ece-gui-007.ece.local.cmu.edu

RHEL7

ssh, fastx

NA

12

64

48

Data Science Cloud (DSC)

Interested ECE community members should visit the access request page for more information.

System Group

Access Methods

Processor

GPU

Total Cores

Memory (GB)

Scratch Storage (TB)

dsc-01.ece.local.cmu.edu

ssh, fastx

 

 

4 x Intel Xeon Processor E7-8870 v3 2.1GHz

NA

72

3,072

14

dsc-02.ece.local.cmu.edu

ssh, fastx

2x Intel® Xeon® Processor E5-2698 v3 2.3GHz

2 x NVIDIA Tesla K40M

36

768

1.8

 

Related pages