ARCC Masthead Image

ARCC Services: Supporting UW Research

 

 

How to Reach Us:

E-mail: arcc-help@uwyo.edu
Request our Services: on our Service Portal
Virtual Office Hours 2x a week via Zoom:
11am-1pm Tuesdays and 1:30-3:30pm Wednesdays
Phone: (307)-766-3600

Mailing Address:
UW Advanced Research Computing Center
Office of Research & Economic Development
Department 3355
1000 E. University Ave
Laramie, WY 82071

About Us:

UW ARCC is the primary research computing facility for the University of Wyoming and our department is housed within the Division of Research and Economic Development.  Our expert staff are committed to providing the UW research community with access to specialized research computing infrastructure, knowledge, support, and a large range of scientific software pre-configured for your use. If you are new to ARCC, please see our getting started guide.   

We manage and maintain and support all internally housed scientific computing resources including HPC and high performance data storage.  UW ARCC also aims to support all high performance computing resources available to any UW researcher.  This includes the support of both UW ARCC hosted HPC resources and External HPC Resources.

A list of our internally offered services is detailed below.

 

 

 

 

 

 

  • Ethernet Cables coming out of Nodes

    High Performance Computing

    We host, maintain and support several in-house HPC resources for use by the UW research community and collaborators:
    • MedicineBow
      • The MedicineBow Compute Environment (AKA MedicineBow) is ARCC's newest HPC cluster.  Released to UW July of 2024, MedicineBow has enhanced GPU offerings to enable research in key areas including AI, machine and deep learning, and complex modeling. 
        • The cluster currently consists 25 96-core CPU nodes, and partitions consisting of A30, L40S, and SXM5 H100 GPU nodes. 
        • The MedicineBow Hardware Summary table provides detailed information about the hardware makeup of this cluster.
      • MedicineBow HPC operates under the condo paradigm
        • Operating within the condo structure, MedicineBow computational resources are often available to UW researchers at no cost.
        • Beyond standard access to the cluster, UW researchers can choose to invest in hardware to gain priority access.  
      • OnDemand offers easy-to-use portal to Medicinebow, providing user's access to the cluster through a web browser)
      • Given the wide set of use cases supported by the MedicineBow cluster, both the Beartooth and WildIris clusters will be consolidated into MedicineBow by EOY 2024.
    • Beartooth
      • The Beartooth Compute Environment (AKA Beartooth) is ARCC's prior generation HPC cluster with roughly ~300 compute nodes and 1.2 PB of storage. 
        • By the end of 2024, most Beartooth nodes will be consolidated into MedicineBow and the Beartooth HPC system will be decommissioned.  
    • Loren 
      • A Specialty GPU based HPC cluster for exclusive use by the Center of Innovation for Flow through Porous Media Research Group
  • Racks down a hall in a datacenter

    High-Performance Data Storage

    In a world of data abundance, research is an increasingly data-intensive activity. We provide data storage as a core service:
    • ARCC Data Portal
      • Our newest high performance data storage implementing state-of-the-art data deduplication using VAST Data technology.  The Data Portal provides over 3PB of NVMe storage.  Researchers can utilize ARCC data storage on this platform by requesting Alcova research storage allocations, or when using ARCC HPC cluster mounted storage.
    • Alcova  (High performance data storage)
      • Designed as a highly collaborative space geared toward project-oriented data storage for use by UWYO research groups and their collaborators. 
      • Allocations on Alcova allow researchers to reliably store and exchange data with anyone located anywhere in the world using Globus 
      • Extremely high transfer speeds.
      • Easily accessible from Windows (mapped drive) and Mac (cifs/smb) workstations.
      • 20 day snapshots for recent accidental data loss recovery.
      • Daily backups
    • Pathfinder (Cloud-like, low-cost data storage)
      • A low-cost storage solution enabling a cloud-like presence for research data hosted by UW ARCC.
      • Designed for scaling and expansion
      • High level data protection. 
      • Core functionality also hosts onsite backups and enables data sharing and collaboration.
      • Utilized for publication of research data in partnership with UW Libraries.
      • Time-based token-authenticated links allow for file sharing through an easily accessible URL.
  • Woman providing support and tutoring

    User Support & Misc Services

    We work to support all research computing at UW as well as specialty resources falling outside our current services:
    • Software Services
      • Consultation for installs and troubleshooting software on HPC platforms.
      • A wide variety of documentation on our wiki to assist users with specialized research software.
    • Training and Workshops
      • We host several trainings and workshops for commonly requested topics.
      • We also support and assist you while hosting a training on our HPC resources.  
    • General User Support
      • Comprehensive Wiki with useful pages and answers common questions.
      • Support staff are regularly available 2x a week over zoom.  
      • Individual Consultations upon request
    • Linux Desktop Support
      • UW ARCC provides campus licensing for Red hat and support for RedHat and Ubuntu.  Users may request ARCC's assistance for specific Linux distributions.
    • HPC Project Migration
      • UW ARCC can provide assistance migrating project data and code.
    • ARCC Investment Program
      • Inquiries may be made through our service portal
      • Node purchase estimates are available upon request.
    • GitLab A collaborative code development service hosted on ARCC resources
    • Globus (Cloud-based saas for data sharing)
      • Used for fast and reliable transfer, sharing, syncing, and publishing of large amounts of research data across HPC platforms and devices.