Frederick Research Computing Environment

Services

A range of services is generally available on the FRCE cluster to best support the science performed at the NCI and the FNLCR. This is not an exhaustive list and users are encouraged to contact the HPC team at any time  to discuss their particular needs.

  1. Batch Processing - The batch processing service provides access to the specialized FNLCR Linux HPC cluster. The clustered nodes provide a uniform file system layout with shared and private storage and centralized access to high-performance servers. Jobs are scheduled through the Slurm workload manager based on current usage of the cluster, the resource requirements of each individual job, and the job history of the submitter. Jobs can range from short single-processor jobs to ones requiring multiple cores for several weeks, and are executed in parallel with other jobs on high-density GPU nodes. Fast-access scratch storage further enhances job execution.

  2. Scientific Software Support - The FRCE infrastructure supports computational needs relating to bioinformatics, computational chemistry, gene sequencing, image analysis, including multiple forms of microscopy, and statistical computing and graphics. Expansion efforts are focused on a highly flexible and leading-edge infrastructure to effectively support customized activities for the science community. A large number of primarily open-source scientific software is installed on the FRCE cluster and additional packages will be installed upon request. A database of installed applications is maintained by ABCS and is available on their web site.

  3. Data Management Consulting - The FRCE infrastructure has access to multiple tiers of storage ranging from high-speed shared flash storage to long-term archival storage. Each user is allocated storage in a home directory and has access to both local and shared scratch storage. Additional long-term storage is available by request and the EIT HPC and Storage teams will work with individual users to help optimize workflows and data transfers.

  4. Workflow Consultation - FRCE infrastructure staff work with the user community to understand job workflow and sequencing needs and then provide recommendations on best resource utilization of the FRCE capabilities.

  5. Job Monitoring - Slurm and XDMoD, a metrics tool used for resource tracking in high-performance computing environments, together provide a comprehensive reporting profile to the user community and operations staff pertaining to resource utilization and job performance and status. Other user and system metrics are collected on a monthly basis and can be made available on request.

  6. Account Creation and Maintenance - any person within the NCI with a need for the computational environment provided by FRCE is encouraged to request an account.

Workloads supported on FRCE include:

  • Molecular dynamics modeling/simulation
  • Cryo-EM
  • Genomics/High Throughput Sequencing
  • Linkage/Phylogenetics
  • Mass Spec
  • Systems Biology
  • Structural Biology
  • Image Analysis
  • Machine/Deep Learning
  • Computational Chemistry
  • XRay/NMR