The series of Lichtenberg high-performance computers at TU Darmstadt provides 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.
In mid-July, Lichtenberg II high-performance computer of TU Darmstadt was officially put into operation. Equipped with the latest technology, it sets standards in performance and energy efficiency and thus offers the best conditions for excellent research. Our two-minute film provides insights into the ceremonial inauguration and a tour of the high-performance computer.
In 2020, the first expansion stage of the Lichtenberg II system with 643 computing nodes was put into operation; this will be expanded by 581 computing nodes with the second expansion stage. Together, the two expansion stages will provide a theoretical peak performance of approximately 8.5 petaflops per second (PFlop/s) through processors and 1.7 petaflops per second through accelerators. The main memory totals 563 terabytes, the storage system for data around 6 petabytes.
Energy-efficient computer systems and sustainable use are key goals for TU Darmstadt. For this reason, the waste heat from Lichtenberg II is not simply released into the environment, but during the heating period, a significant portion is fed into the district heating network that connects all buildings on the Lichtwiese campus.
For this purpose, Lichtenberg II uses direct and highly efficient hot water cooling to fully utilize the power of the processors. In the process, special heat exchangers and coolant distributors enable high return temperatures of more than 45 degrees Celsius to ensure sensible reuse of energy and efficient cooling. This leads to a significantly improved CO2 and energy balance and is an important step towards sustainable high-performance computing.
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.
The experts of the (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. “Hessisches Komptenzzentrum für Hochleistungsrechnen”
Power Rail Short & Fire L5|08 [Update 2025-03-26]
March 25, 2025
HPC & Housing are inoperative
Due to a short and subsequently, a fire in the power rail system, the TURZ-L5|08 data center is currently inop, affecting the Lichtenberg HPC, network and housing.
Preparations for Migration to RedHat EL 9
February 24, 2025
Jumping up a major release of the cluster's operating system
Some login and compute nodes are already migrated to RHEL 9.4
Solved: Failure of the Cluster-wide File System
November 04, 2024
System back at normal and available
+++ Update 17:00: The deadlock problem could only be fixed by (hard) reset of various GPFS master servers and a reboot of all compute nodes. Hence, all running jobs at the time of the GPFS lockup unfortunately are lost. If you did not explicitly prohibit it (by using special parameters), the scheduler will restart those jobs on its own. +++