Workstations hosted at CERN
The following workstations are avaiable for development and production. - Access is granted based on the use case. - For further information and to get access please contact abp.computing[AT]cern.ch.
|pcbe-abp-gpu001||AMD Ryzen Threadripper 2970WX 24-Core||128 GB||4x Titan V|
|pcbe-abp-hpc001||AMD Ryzen Threadripper 3990X 64-Core||256 GB||GeForce GTX 1660 Ti|
|liupsgpu||2x Intel Xeon E5-2630 6-core||256 GB||3x Tesla C2075|
When possible the CERN batch system (HTCondor) should be used for production studies while these workstation should be used for special cases (e.g. development, performance benchmarking, computations requiring very large amount of memory etc.).
- A user can cange the password py simply typing
- An administrator can add a user by:
(the -m flag creates the home directory).
sudo adduser username
- An administrator can change the password for a user by typing:
sudo passwd username
Workstations hosted at INFN-CNAF
GPUs are available also at CNAF (for users having a CNAF HPC account - more info available here). You can launch GPU jobs interactively with the following steps:
- Connect with SSH into the login server:
- Connect wiht SSH to the node that has the 4x Nvidia Tesla V100 GPUs:
- Enable newer versions of gcc, cmake etc. with
scl enable devtoolset-7 bash