Skip to content

Cluster Description

The XLence cluster is a high-performance computing facility for scientific research in structural bioinformatics, NGS data analysis, and proteomics.

Hardware Resources

Login Node

The cluster has 1 login node with redundant backup for high availability.

  • Users connect to this node via SSH to access the cluster
  • Used for job submission, data management, and light file operations
  • In case of hardware failure, service can be restored within hours to a few days

⚠️ Important: The login node is for interactive work and job submission only. Computationally intensive tasks must be submitted to the compute nodes via the job scheduler.

Compute Nodes

General Purpose (5 nodes) - Partition: normal

  • CPU: 10 cores per node
  • RAM: 128 GB per node
  • GPU: 2 GPUs per node

These nodes are optimized for structural bioinformatics, molecular dynamics, and GPU-accelerated workloads.

NGS Analysis (1 node) - Partition: ngs

  • CPU: 18 cores
  • RAM: 128 GB
  • GPU: 1 GPU

This node is specialized for Next Generation Sequencing (NGS) analysis and memory-intensive workloads.

Shared Storage

All cluster nodes share access to centralized storage:

  • General storage: 50 TB
  • NGS data storage: 10 TB (reserved for sequencing data)

Your home directory and data are accessible from all nodes in the cluster.

Job Scheduling

The cluster uses Slurm as the workload manager for job submission and resource allocation.

Running Computations

All computationally intensive tasks must be submitted to the compute nodes using Slurm. This includes:

  • Molecular dynamics simulations
  • Structure predictions
  • NGS data processing
  • Large-scale data analysis
  • Any long-running calculations

⚠️ Login Node Policy: Programs found running heavy computations on the login node will be terminated by the administrators without notice.

Submitting Jobs

To run your computations on the compute nodes, use the Slurm job scheduler.

For detailed instructions, see: Job Scheduling with Slurm

Available Partitions

  • normal: General-purpose computing (5 nodes, 10 CPUs + 2 GPUs per node)
  • ngs: NGS data analysis (1 node, 18 CPUs + 1 GPU)

Support

For questions about cluster resources or job scheduling:

Last Updated: October 2025