NEC Cluster (Vulcan)

Image f. NEC Cluster (Vulcan)

Initially installed in 2009, Vulcan has been continually upgraded to address new needs of our system users. The configuration of this standard PC cluster is continually adapted to meet increasing demands and provide requirement-optimized solutions, including CPU, GPU, and vector computing components. The current configuration is as follows.

System information

Intel Xeon Gold 6248 @2.5GHz (CascadeLake)

Number of nodes: 96
Memory per node: 128 GB

Intel Xeon Gold 6138 @2.0GHz (SkyLake)

Number of nodes: 100
Memory per node: 192 GB

AMD Radeon

CPU: Intel Xeon Silver 4112 @ 2.6 GHz (Skylake)
Number of nodes: 6
Memory per node: 96 GB
GPU: 1 x AMD Radeon Pro WX8200
GPU memory: 8 GB

AMD Genoa

CPU: AMD Genoa Epyc 9334 @ 2.7GHz base, 32 cores
Number of nodes: 156
Memory per node: 768 GB

Nvidia A30

Number of nodes: 24
CPU: 2 * AMD Epyc 9124 Genoa, 3.0GHz base, 16 core
Memory per node: 768GB DDR5
GPU: 1 x Nvidia A30
GPU memory: 24GB HBM2e

Large Memory Nodes

Number of nodes: 2
CPU: 2 * AMD Epyc 9124 Genoa, 3.0GHz base, 16 core
Memory per node: 3072GB DDR5

Are you a scientist or engineer interested in using HLRS's computing resources? Learn how to get access to our systems and expertise.

Applying for computing time

NEC SX-Aurora TSUBASA A300-8 @ 2.6 GHz

Number of nodes: 8
Memory per node: 192 GB
Vector engines: 8 x NEC Type 10B @ 1.4 GHz
Vector engine memory: 48 GB @ 1.2 TB/second

Interconnects

Infiniband EDR/HDR

Related news

All news

German National Supercomputers Combined to Run a Single Large-Scale Simulation

Profile: A Life in Supercomputing

HLRS Wins Datacenter Strategy Award in the Category "Transformation"

HLRS Completes TISAX Level 3 Assessment for Information Security