Computing Infrastructure

The Computing Infrastructure offered by IT to a (more or less) general DESY public - and beyond.

Publc Linux login and workgroupserver

IT offers gerneral purpose Linux login nodes for casual, small scale usage in the DESY intranet.

More information about the PAL service can be found here

Some groups operate their own login systems, please contact your group administrator if necessary.

 

Public Windows Terminal Service

IT offers gerneral purpose Windows terminal serivce for casual, small scale usage in the DESY intranet.

More information about winterm can be found here

Due to licensing issues, not all users from all groups have access. Please contact UCO or your group administrator for details.

Some groups operate their own Windows terminal service, please contact your group administrator if necessary.

 

Maxwell: High Performance Compute Infrastruktur

The Maxwell HPC system was developped with a special focus on the analysis of data from XFEL, Petra-III and FLASH, the simulation of new experiments and acceleratores, and the usage of artificial intelligence algorithms.

The Maxwell HPC system offers interactive access via login nodes and Jupyter notebooks, and features a large batch system with about 750 servers (status mid 2022). A dedicated high performance network enables both massive parallel computations and fast data access.

More information, especially about access to the system, can be found on the documentation page.

 

NAF (with BIRD): National Analysis Facility

The NAF was designed for the analysis and simluation of data from the particle physics experiments ATLAS, CMS, BELLE and ILC. Meanwhile, also a large number of smaller groups and experiments are successfully using the NAF.

The NAF offers interactive access via login nodes and Jupyter notebooks, and features a large batch system with aboyt 350 server (status mid 2022). The NAF is optimized for interactive and fast turn-around analysis of large data sets, when job profiles do not need inter process communication.

The batch system previously calles BIRD has been fully integrated in the NAF.

More information, especially about access to the system, can be found on the general documentation page , and information about batch usage more specifically on  this dedicated page zu finden

 

GRID: Worldwide accessible compute infrastructure

Since almost two decades, the Grid is the working horse for lage scale analyses and simluations of data for particle physics experiments.

First developped in the context of the LHC experiments, DESY versy soon utilized grid technologies also for their own experiments such as then-active HERA, but also ILC and BELLE. In the context of the LHC Computing Grid, DESY acts as a Tier-2 center.

The DESY Grid infrastructure offers a large batch system with about 400 servers (status mid 2022), equipped with standardized, worldwide accessible interfaces. In addition, DESY is one of the few centers worldwide to also offer further services, that enable the usage of the worldwide distributed Grid infrastructure for some experiments.

More information can be found on the  documentation page.