Toulouse:Hardware

From Grid5000
Revision as of 10:48, 15 September 2023 by Pneyron (talk | contribs)
Jump to navigation Jump to search

See also: Network topology for Toulouse

Summary

  • 2 clusters
  • 21 nodes
  • 384 CPU cores
  • 12 GPUs
  • 2.63 TiB RAM
  • 21 SSDs and 0 HDDs on nodes (total: 32.65 TB)
  • 33.1 TFLOPS (excluding GPUs)

Clusters

Cluster Access Condition Date of arrival Nodes CPU Memory Storage Network Accelerators
# Name Cores Architecture
estats testing queue,
exotic job type
2023-06-13 12 1: 1
[2-12]: 4
Carmel Unknown 1: 8 cores/CPU
[2-12]: 2 cores/CPU
aarch64 32 GiB 2.0 TB SSD 1 Gbps  Nvidia AGX Xavier (30 GiB)
montcalm testing queue 2022-12-01 9 2 Intel Xeon Silver 4314 16 cores/CPU x86_64 256 GiB 960 GB SSD 10 Gbps 

**: crossed GPUs are not supported by Grid'5000 default environments

Clusters in the testing queue

estats

12 nodes, 45 cpus, 96 cores, split as follows due to differences between nodes (json)

Reservation example:

Terminal.png ftoulouse:
oarsub -q testing -t exotic -p estats -I
estats-1 (1 node, 1 cpu, 8 cores)
Access condition: testing queue, exotic job type
Model: Connecttech/Nvidia Jetson AGX Xavier
Date of arrival: 2023-06-13
CPU: Carmel Unknown (Carmel, 1 CPU/node, 8 cores/CPU)
Memory: 32 GiB
Storage: disk0, 2.0 TB SSD NVME Samsung Samsung SSD 970 EVO Plus 2TB (dev: /dev/disk0, by-path: /dev/disk/by-path/platform-14180000.pcie-pci-0000:01:00.0-nvme-1) (primary disk)
Network:
  • eth0, Ethernet, configured rate: 1 Gbps, model: N/A, driver: nvethernet - no KaVLAN
GPU: Nvidia AGX Xavier (30 GiB)
Compute capability: 7.2

estats-[2-12] (11 nodes, 44 cpus, 88 cores)
Access condition: testing queue, exotic job type
Model: Connecttech/Nvidia Jetson AGX Xavier
Date of arrival: 2023-06-13
CPU: Carmel Unknown (Carmel, 4 CPUs/node, 2 cores/CPU)
Memory: 32 GiB
Storage: disk0, 2.0 TB SSD NVME Samsung Samsung SSD 970 EVO Plus 2TB (dev: /dev/disk0, by-path: /dev/disk/by-path/platform-14180000.pcie-pci-0000:01:00.0-nvme-1) (primary disk)
Network:
  • eth0, Ethernet, configured rate: 1 Gbps, model: N/A, driver: nvethernet - no KaVLAN
GPU: Nvidia AGX Xavier (30 GiB)
Compute capability: 7.2

montcalm

9 nodes, 18 cpus, 288 cores (json)

Reservation example:

Terminal.png ftoulouse:
oarsub -q testing -p montcalm -I
Access condition: testing queue
Model: HPE Proliant DL360 Gen10+
Date of arrival: 2022-12-01
CPU: Intel Xeon Silver 4314 (Ice Lake, 2.40GHz, 2 CPUs/node, 16 cores/CPU)
Memory: 256 GiB
Storage: disk0, 960 GB SSD SATA HP VK000960GXAWL (dev: /dev/disk0, by-path: /dev/disk/by-path/pci-0000:00:17.0-ata-1) (primary disk)
Network:
  • eth0/ens10f0np0, Ethernet, configured rate: 10 Gbps, model: Broadcom Inc. and subsidiaries BCM57414 NetXtreme-E 10Gb/25Gb RDMA Ethernet Controller, driver: bnxt_en
  • eth1/ens10f1np1, Ethernet, model: Broadcom Inc. and subsidiaries BCM57414 NetXtreme-E 10Gb/25Gb RDMA Ethernet Controller, driver: bnxt_en - unavailable for experiment

Last generated from the Grid'5000 Reference API on 2023-09-15 (commit 055cf2c4f7)