Page tree
Skip to end of metadata
Go to start of metadata

Access

Overview of HPC Systems


ArchitectureTotal NumbersMax Job LimitsAccess
System NameCPUCores per NodeRAM per Node [GB]NodesCoresNodesCoresWall TimeQueues/ PartitionsLogin Node
SuperMUC-NG
Thin Nodes
Intel Xeon ("Skylake")48966336304128316815206448hmicro, general, large,skx.supermuc.lrz.de
SuperMUC-NG
Fat Nodes
Intel Xeon ("Skylake")48768144891212876848hfatskx.supermuc.lrz.de
SuperMUC-NG
Cloud
Intel Xeon ("Skylake")48963215368384tbd

SuperMUC-NG CloudIntel Xeon ("Skylake") + 2x Nvidia V10048963215368384tbd

SuperMUC Phase 2 Haswell NodesIntel Xeon E5-2697 v3 ("Haswell")28643072860165121433648htest, micro, generalhw.supermuc.lrz.de
SuperMUC Phase 2
Fat Nodes
Intel Xeon E5-2697 v3 ("Haswell")28256822482246hbighw.supermuc.lrz.de
SuperMUC Phase 2 Remote VisualizationIntel Xeon E5-2697 v3 ("Haswell") + NVidia GPU K4028256384 + GPU12848hvishw.supermuc.lrz.de, submit job
Linux Cluster CoolMUC-2Intel Xeon E5-2690 v3 ("Haswell")28643841075260168048h96hmpp2, seriallxlogin[5-7].lrz.de,
Linux Cluster CoolMUC-3Intel Xeon Phi (Knights Landing) 64 96 148 9472 25 1600 48hmpp3lxlogin8.lrz.de
Linux-Cluster IvyMUCIntel Xeon E5-2650 ("Ivy Bridge") 1664314961219272hivymuclxlogin10.lrz.de
Linux Cluster TeramemIntel Xeon E7-8890 v4 ("Broadwell") 96 6144 1 96 1 96 48hinterlxlogin[5-7].lrz.de
Machine Learning
Systems DGX-1 and DGX-1v
Nvidia Pascal P100 or V100 8 GPU 128 2 n.a. 1 n.a. 12h
contact servicedesk, www.datalab.lrz.de
GPU CloudNvidia Pascal P100 1 GPU 128 4 n.a. 1 n.a. 12h
contact servicedesk, www.datalab.lrz.de
LRZ Compute CloudIntel Xeon E5540, X5650, E5-2660v21-201-24095852
64
N.A.www.cloud.mwn.de
Self-Service VMs with custom operating system
LRZ Virtual MachinesIntel Xeon E5-2660 v2 ("Sandy Bridge")1-81-32901800


Managed Hostinghttp://www.lrz.de/services /serverbetrieb/




  • No labels