Power your research with our computing infrastructure

Research Computing (RC) offers the O2 cluster, a dynamic high-performance computing (HPC) environment that supports a broad spectrum of research projects. From the latest sequencing analysis and molecular dynamics to mathematical modeling and image analysis, O2 is built to handle diverse computational requirements. Whether you're running intensive tasks or light computations, the O2 cluster is designed to deliver optimal performance. We provide bioinformatics consulting and other specialized assistance to support your research further.

To sign up to use HMS HPC services:

  • Eligibility

    HMS Quad-based labs and HMS affiliate hospital and institution labs can access our HPC services
    Researchers Affiliates

    Labs whose principal investigator has a primary or secondary faculty appointment in an HMS Blavatnik Institute Basic or Social Science Department will not be charged for using O2 or storage resources.

    Labs whose PIs do not have a primary or secondary faculty appointment in an HMS Blavatnik Institute Basic or Social Science Department will be charged for using HMS High-Performance Computing and storage resources. For information about the RC Core and charges, please see the Service rates selection below.

  • Security

    Your data's security is our top priority
    Encryption Backup

    Research Computing ensures stringent security measures. Data stored and processed through O2 is encrypted, and regular backups ensure data integrity and availability.

  • Features

    Cutting-edge infrastructure and tools to power your research
    Hardware Software

    The O2 cluster boasts over 11,000 compute cores, 350 nodes with 256 GB RAM each, and nine high-memory nodes with up to 1 TiB RAM. Read more about our hardware.

    With over 170 applications and tools for computational analysis, O2 offers extensive software options.

  • Service rates

    Billing rates by type of computing and storage resource.
    Compute GPU

    Compute

    The compute charges depend on the resources you request. Costs are determined by total compute hours in a given billing period.

    • Standard Compute – Jobs with less than 256GiB RAM reserved are billed at $0.0105 per core hour and $0.0013 per GiB RAM hour.
    • High Memory Compute – Jobs with more than 256GiB RAM reserved are billed at $0.0144 per core hour.

     

    GPU

    • GPU Compute – Single or double precision tasks are billed at $0.3428 per card hour.

     

    Storage

    Storage charges are computed based on the sum of daily usage over the days in a quarter.

    • Active Compute – Fast SSD for HPC Compute (O2) at $224.06 per TiB annually or $56.02 quarterly.
    • Active Collaborations – Disk for lab computers at $169.65 per TiB annually or $42.42 quarterly.
    • Standby – For infrequently accessed data at $111.64 per TiB annually or $27.91 quarterly.

     

    Annual fees

    The annual fee is for non-quad-based HMS groups with a group folder or active cluster accounts.

    • Annual Access Fee – $40.00 per principal investigator or group.

    Billing rates are reviewed annually.

Comprehensive guide to our high-performance computing services.

Web interface to access the O2 cluster and its tools.

Courses on the O2 HPC cluster, Python, MATLAB, and more. Open to all HMS affiliates.

Weekly consultation sessions are held every Wednesday from 1–3 PM via Zoom.