High Performance Computing Group

Juan Carlos Perin (Director)




Visit the HPC’s Intranet site

More »

The MSK HPC Group is home to powerful computing systems designed to facilitate the work of the Computational Biology Center and any MSK/SKI researchers, in such areas as:

  • high-throughput DNA and RNA sequencing
  • proteomics
  • protein and drug interaction modeling
  • denovo assembly and structural variation
  • data mining and modeling
  • high-resolution imaging/image analysis

We currently operate five distinct HPC clusters. All are Linux compute clusters, and each is composed of multiple compute nodes/servers linked with 10GB networking and attached via fast connections to several data storage appliances providing an aggregate of about 11 Petabytes of storage for our users. 

Our primary research HPC cluster is known as “Hal.” Its configured with 3 head nodes, and over 2100 compute cores/processors (CPUs) with a minimum of 2-8 GB of RAM memory per core.  Many of these systems are also equipped with GPU computing options from NVIDIA. These systems use the Torque/Moab queuing system, which allow users to submit complex jobs for the cluster to compute in parallel from a terminal with SSH access.

The other HPC systems are dedicated to various tasks such as genome sequence primary analysis, clinical diagnostics sequence analysis, bioinformatics pipeline work, and other compute intensive tasks that we attempt to keep separate from applied research work.

Our High Performance Computing Group provides MSK researchers with access to high-performance computing (HPC) resources including machines (computer hardware), analytical tools, storage space, and technical support for solving data-intensive research problems.