Computing power < Superkomputer

Kraken (Fregata and Tryton Plus)

 

TFLOPS

 

Kraken - Fregata

Cluster purchased as part of the PRACE-LAB2 project.

CPU Partition
Platform HPE Cray XD220v
Processors Xeon Platinum 2.0 GHz, 56-core
Memory 512 GB per server
Number of nodes: 80
GPU Partition
Platform HPE Cray XD665
Processors AMD EPYC 9334 2.7 GHz, 32-core
GPU Nvidia HGX H100-94 SXM5 x4
Memory 768 GB per server
Number of nodes 37
General Information
Network InfiniBand NDR200 fat tree topology
Operating System Linux
Computational Power 10.6 PFLOPS
Additional Equipment Supporting service servers, SDN network management system cloud servers, 10/40/100 Gbps Ethernet Fabric network switch team, CDU direct liquid cooling exchangers (2 cabinets)

 

Kraken - Tryton Plus

Cluster purchased in mid-2021 as part of the PRACE-LAB project.

CPU Partition
Platform Huawei CH121L V5 Liquid-Cooled Compute Node
Processors Xeon Platinum 2.9 GHz, 24-core
Memory 192/384 GB per server
Total 634 servers, 1268 processors, 30432 cores, 166 TB RAM
Cabinets 18 including 3 CDUs
GPU Partition
Platform HPE Cray XD665
Processors AMD EPYC 9334 2.7 GHz, 32-core
GPU Nvidia HGX H100-94 SXM5 x4
Memory 256 GB per server
Number of nodes 16
General Information
Network InfiniBand Fabric EDR 100 Gbps, fat tree topology
Operating System Linux
Computational Power 3.8 PFLOPS
Additional Equipment Supporting service servers SDN network management system, OpenStack Ironic provisioning system, 10/40/100 Gbps Ethernet Fabric network switch team, CDU direct liquid cooling exchangers (3 cabinets)

 


Roro 2

Storage device cluster. Used for providing resources for Tryton/KDM and Tryton/Plgrid:

  • Workspace (Lustre)
  • Home directories (NFS)
Type storage device cluster
Name roro2
Servers
Device ACTINA SOLAR 220 X5
Processors 2x Intel Xeon 8C E5-2650
RAM 64GB DDR3 1600 MHz ECC
Storage 2x Intel® SSD DC S3700 Series 100GB
Network Mellanox ConnectX-2  Dual Port QSFP IB 40Gb/s
Quantity 8
Metadata and NFS Disk Array
Device Hitachi Unified Storage 150
Cache Memory 2 x 16GB
Disk Memory 150 x 900GB/10K SAS-2 2.5"
Network 8 x FC 8Gb
Quantity 1
Disk Array for Workspaces
Device Hitachi Unified Storage VM
Cache Memory 2x 128GB
Disk Memory 184 x 3TB/7.2K SAS-2 3.5
Network 16 x FC 8Gb
Quantity 4
   
Installation/Start Date 25.02.2014
Upgrade History

2015: Expansion under Plgrid project: 3 servers, one workspace matrix, 211TB additional disks for metadata matrix

Integrator  WASKO S.A.

 

Retired ->

CONTACT

Our consultants help future and novice users of specialized software installed on High Performance Computers (KDM) at the TASK IT Center.

Contact for High Performance Computers, software / licenses, computing grants, reports:

kdm@task.gda.pl

Administrators reply to e-mails on working days between 8:00 am – 3:00 pm.