Configuration

Hardware Configuration of the JUSUF Cluster Module

  • 144 standard compute nodes
    • 2× AMD EPYC 7742, 2× 64 cores, 2.25 GHz

    • 256 (16× 16) GB DDR4, 3200 MHz

    • InfiniBand HDR100 (Connect-X6)

    • local disk for operating system (1× 240 GB SSD)

    • 1 TB NVMe

  • 61 accelerated compute nodes
    • 2× AMD EPYC 7742, 2× 64 cores, 2.25 GHz

    • 256 (16× 16) GB DDR4, 3200 MHz

    • InfiniBand HDR100 (Connect-X6)

    • local disk for operating system (1× 240 GB SSD)

    • 1 TB NVMe

    • 1× NVIDIA V100 GPU with 16 GB HBM2e

  • 4 login nodes
    • 2× AMD EPYC 7742, 2× 64 cores, 2.25 GHz

    • 256 (16× 16) GB DDR4, 3200 MHz

    • HDR 100-Infiniband (Connect-X6)

    • 100 Gigabit Ethernet external connection

    • local disk for operating system (2× 480 GB SSD in RAID 1)

  • 26,240 CPU cores

  • 944.6 (CPU) + 427 (GPU) Teraflop per second peak performance

  • Mellanox InfiniBand HDR full fat-tree network with HDR100 speed on the nodes and full HDR on the inter-switch level

  • 100 Gbit/s network connection per login node and 40 Gbit/s network connection per compute node to JUST for storage access

Software Overview

  • CentOS 7 Linux distribution

  • Parastation Cluster Management

  • ParTec ParaStation MPI (Message Passing Interface) implementation

  • Intel MPI

  • Open MPI

  • IBM Spectrum Scale (GPFS) parallel file system