Lichtenberg High Performance Computer of TU Darmstadt
The series of Lichtenberg high-performance computers provide computing resources for researchers from academia and public research facilities in Germany.
The multifaceted architecture of the high performance computer allows for flexible and efficient scientific computing, especially for computationally intensive applications. The experts of the “Hessischen Komptenzzentrum für Hochleistungsrechnen” (Hessian Competence Center for High Performance Computing) will provide for consultancy and scientific support on using the cluster efficiently and hosts various workshops on scientific HPC. The new high-performance computer Lichtenberg II includes large memory nodes (this time with 1.5 TByte RAM), as well as new “accelerator” systems with dedicated NVIDIA GP-GPUs. Further, an expansion in the form of several “DGX A100” systems is currently underway, which will support modern AI research.
The first expansion stage (including this year's expansion) of Lichtenberg II will in future provide its users with a real peak performance of over 3,148 PFlop/s and a total of 257 TByte RAM. This also includes a new high-performance storage system with around 4 PByte for data, most of which has already been operational in 2019. By the end of 2021, the Lichtenberg II high-performance computer will be finally extended and completed by its second major expansion stage.
Furthermore, the new cluster expansion stage not only offers high computing power, but is also operated very energy-efficiently, – see Performance and efficiency.
The entire Lichtenberg II system (all expansion stages) is funded by the German Federal Government and the State of Hesse with a total amount of ~15 million euros.
Interested users can register here (see Access). The use of this service requires a application, which is reviewed by the Research Computer Advisory Board of the TU Darmstadt. You can find introductory explanations on how to use the cluster, such as login nodes and the queuing system, under Using the Cluster. Information on hard- and software can be found under Operations.
Attendance is free of charge
For the (potential) users of the Lichtenberg supercomputer, a monthly introduction will be held online. Subjects are the available hardware and software and the general use of the (batch) system. go
Refurbishments on the cooling systems
Over the week from 4th – 10th. October, the whole HPC cluster will be down and unavailable. go
As you might know, TU Darmstadt is one of the national HPC centers (NHR) since the beginning of this year . RWTH Aachen University and the TU Darmstadt are together forming NHR4CES  to cooperate in the context of NHR . This advances HPC nationwide and secures funding for the next decade. However, this also means, that there will be changes in the way that compute resources are granted and applied for. Regarding Tier 2 project categories („Medium“ and „Large“), we do not differentiate between members from TU Darmstadt and other members of German public or government-approved teaching and research institutions anymore. This means that the project categories “Medium” and “Large” will not be available for new applications anymore. Currently running projects will continue as scheduled. “Small” Projects can be applied for as before. go