Nieuwpoort cluster

Request account

The Nieuwpoort cluster is accessible to all members of the Berendsen Center groups. Please send an email to your PI to request for an account

Courses

The CIT offers a beginners course for using the Habrok cluster. The Habrok cluster in priciple works the same as the Nieuwpoort cluster. More information can be found on the course page.

Interactive jobs

Interactive jobs and visualization software should be run on the "interactive" node (not the head node). This includes compiling of larger programs, generating movies with VMD, etc. You can access this node with 'ssh interactive'.

The cluster consist of different compute node types

• 40 Standard Nodes (node001 to node040)
  • 2 CPUs (14 Cores each/28 Threads)
  • 1 TB Disk
  • 128 GB ram
• 2 GPU Nodes (gpu01 to gpu02)
  • 2 GPUs NVIDIA TESLA K40M
  • 2 CPUs (14 Cores each/28 Threads)
  • 1 TB Disk
  • 128 GB ram
• 1 High Memory Node (himem01)
  • 2 CPUs (14 Cores each/28 Threads)
  • 1 TB Disk
  • 768 GB ram
• 1 Large Disk Node (ldisk01)
  • 2 CPUs (14 Cores each/28 Threads)
  • 2 TB Disk
  • 128 GB ram
• 4 Large Disk & Higher Mem Node (ldisk02 to ldisk05)
  • 2 CPUs (24 Cores each/48 Threads)
  • 1 TB Disk + 4 Tb Scratch disk
  • 192 GB ram
• 1 Interactive node (interactive)
  • 2 CPUs (14 Cores each/28 Threads)
  • 1 TB Disk
  • 128 GB ram

The nodes are coupled using an 56 Gb/s Infiniband network. This network has both a high bandwidth (large amounts of data per second) and a low latency (minimal amount of time needed for sending small messages). This makes this network especially suited for tasks using more than one machine.

Disk space and management

• The drive /users contain 22 TB of space of which 10 GB is assigned to each user.
• The drive /data contain 175 TB of space of which 2 TB is assigned to each user.

The /users drive is intended for location of important files, scripts, and own programs for development. The /data drive is intended for all data files.

It is recommended that you check your disk quota (with the 'quota -vs' command) before submitting any job to make sure that the output fits. You should also regularly check the available disk space with the 'df -h' command to check that the disk systems are not full.

Full disks result in crashing jobs.

When you clean up your directories, you may find the 'du -sh' command convenient as it lets you find out how much space you have in a particular folder specified.

Additional disk space can be assigned to high priority Berendsen Center project upon approval by the center PIs.

System Administration

The system administration is performed by Jaap Aikema (J.L.Aikema@rug.nl).