Skip to main content

Hardware and Software Configuration of Jean Zay

General Information​

Jean Zay is a supercomputer consisting of an HPE SGI 8600 part and a Bull BullSequana XH3000 part, forming a total of five compute partitions: one partition containing scalar nodes (with only CPUs) and four partitions containing accelerated nodes (hybrid nodes equipped with both CPUs and GPUs).

The HPE SGI 8600 compute nodes are interconnected by an Intel Omni-Path network, and those of the Bull BullSequana XH3000 extension are interconnected by an InfiniBand network.

All nodes access a Lustre high-bandwidth parallel file system.

Three storage technologies are available: an SSD disk storage with a capacity of 8.5 PB, a rotating disk storage with a capacity of 44 PB, and a magnetic tape storage with a capacity of 56 PB.

The operating environment is based on RedHat 9.6 (since 17/02/2026) and Slurm 23.11.3 (since 01/10/2024). The available compilers are the Intel and NVIDIA/PGI compilers.

After three successive extensions, the cumulative peak power of Jean Zay has been 125.9 PFlop/s since July 2024.

jean-zay-hardware

Hardware configuration of Jean Zay

Detail of the Jean Zay Partitions​

HPE SGI 8600 part​

Access to the various hardware partitions of the machine depends on the type of job submitted (CPU or GPU) and the Slurm partition requested for its execution (see the detail of the Slurm CPU partitions and the Slurm GPU partitions).

CPU Intel partition (cpu_p1)​

Jean Zay CPU

By default or by requesting the partition cpu_p1, you will have access to the following resources:

  • 720 scalar compute nodes with:

    • 2 Intel Xeon Gold 6248 processors (20 cores at 2.5 GHz), i.e. 40 cores per node
    • 192 GB of memory per node
note

Following the decommissioning of 808 CPU nodes on 05/02/2024, this partition went from 1,528 nodes to 720 nodes.

GPU NVIDIA V100 partition (gpu_p13)​

Jean Zay GPU V100 16GB
Jean Zay GPU V100 32GB

By default or with the constraint v100-16g or v100-32g, you will have access to the following resources:

  • 396 quadri-GPU accelerated compute nodes with:

    • 2 Intel Xeon Gold 6248 processors (20 cores at 2.5 GHz), i.e. 40 cores per node
    • 192 GB of memory per node
    • 126 nodes with 4 NVIDIA Tesla V100 SXM2 16 GB GPUs (with v100-16g)
    • 270 nodes with 4 NVIDIA Tesla V100 SXM2 32 GB GPUs (with v100-32g)
note

Following the decommissioning of 220 quadri-GPU V100 16 GB nodes on 05/02/2024, this partition went from 616 nodes to 396 nodes.

GPU NVIDIA V100 partition (gpu_p2)​

Jean Zay GPU P2

With the gpu_p2, gpu_p2s or gpu_p2l partition, you will have access to the following resources:

  • 31 octo-GPU accelerated compute nodes with:

    • 2 Intel Xeon Gold 6226 processors (12 cores at 2.7 GHz), i.e. 24 cores per node
    • 20 nodes with 384 GB of memory (with gpu_p2 or gpu_p2s)
    • 11 nodes with 768 GB of memory (with gpu_p2 or gpu_p2l)
    • 8 NVIDIA Tesla V100 SXM2 32 GB GPUs

GPU NVIDIA A100 partition (gpu_p5)​

Jean Zay GPU P5

With the gpu_p5 partition (June 2022 extension), you will have access to the following resources:

  • 52 octo-GPU accelerated compute nodes with:

    • 2 AMD Milan EPYC 7543 processors (32 cores at 2.80 GHz), i.e. 64 cores per node
    • 512 GB of memory per node
    • 8 NVIDIA A100 SXM4 80 GB GPUs

Bull BullSequana XH3000 part​

GPU NVIDIA H100 partition (gpu_p6)​

Jean Zay GPU P6

With the gpu_p6 partition (July 2024 extension), you will have access to the following resources:

  • 364 quadri-GPU accelerated compute nodes with:

    • 2 Intel Xeon Platinum 8468 processors (48 cores at 2.10 GHz), i.e. 96 cores per node
    • 512 GB of memory per node
    • 4 NVIDIA H100 SXM5 80 GB GPUs

Other partitions​

Pre and post-processing​

Jean Zay PREPOST

With the prepost partition, you will have access to the following resources:

  • 4 large memory pre and post-processing nodes with:

    • 4 Intel Skylake 6132 processors (12 cores at 3.2 GHz), i.e. 48 cores per node
    • 3 TB of memory per node
    • 1 NVIDIA Tesla V100 GPU
    • a 1.5 TB internal NVMe disk

Visualisation​

Jean Zay VISU

With the visu partition, you will have access to the following resources:

  • 5 scalar visualisation nodes

    • 2 Intel Cascade Lake 6248 processors (20 cores at 2.5 GHz), i.e. 40 cores per node
    • 192 GB of memory per node
    • 1 NVIDIA Quatro P6000 GPU

Compilation​

Jean Zay COMPIL

With the compil partition, you will have access to the following resources:

  • 4 pre and post-processing nodes (see above)

  • 3 Intel compilation nodes

    • 1 Intel(R) Xeon(R) Silver 4114 processor (10 cores at 2.20GHz)
    • 96 GB of memory per node

Jean Zay COMPIL H100

A compil_h100 partition is also available, giving access to processors identical to those of the Eviden H100 partition (gpu_p6):

  • 1 Intel compilation node

    • 2 Intel(R) Xeon(R) Platinum 8468 processors (48 cores at 2.10GHz), i.e. 96 cores per node
    • 256 GB of memory per node

Archiving​

The partition archive uses the same nodes as the prepost partition (see above).

Front-end nodes​

When you connect via SSH to Jean Zay, you arrive on the following resources:

  • 5 front-end nodes
    • 2 Intel Cascade Lake 6248 processors (20 cores at 2.5 GHz), i.e. 40 cores per node
    • 192 GB of memory per node