"The services provided by ARCC are top of the field. I have worked at several other large research universities and the computing resources do not even compare. The availability of large nodes is great. However the support and help provided by the staff really sets them apart. With these resources no one should be intimidated to start using cluster computing as the great staff will help get you going!"
The University of Wyoming has an established commitment to supporting world-class research by maintaining access to resources and up-to-date tools. This mission is shared by UW Advanced Research Computing and we offer a wide range of High Performance Computing services.
High Performance Computing is a main engines driving the advancement of scientific research today. Our resources can help engineers, data scientists, designers, and other researchers solve large, and complex problems in far less time than traditional computing. If you are a member of the UW research community, ARCC can provide you with access to our primary HPC cluster, Beartooth, at no cost. Beartooth is a combination of specialized hardware including a large collection of powerful computers, consolidated as a single system with distributing processing framework. This means Beartooth is configured to handle large amounts of data at high speeds with parallel performance and high availability.
If your research runs into problems that have outgrown your desktop, laptop, or primary workstation, moving your work to HPC may be an ideal solution. Signs you've outgrown your current computer include:
You're getting errors indicating you no longer have enough memory, disk space. This may also result in a non-responsive computer or crashing of your computational processes/software.
Your program needs to be run a large number of times (usually with alternating input data).
It takes too long to run your computations, but they could run faster in parallel (software capable of this will usually have keywords indicating their use with MPI or OpenMP).
Your computations/analysis need access to GPU acceleration.
If any of these situations seem familiar, you can likely benefit from moving your research to HPC. Click here to request access to our HPC cluster through our service portal or e-mail us at arcc-help@uwyo.edu with any questions.
UW ARCC also offers research centered data storage options, a variety of research support software and tools, systems and end user support, and miscellaneous services to help you get your research done in a more efficient and effective way. Click here for a more comprehensive list of our services.
Beartooth is UW ARCC's central High Performance Computing cluster dedicated to research computing at the University of Wyoming. Beartooth is a shared service available to the entire UW research community, and now newly available to private industry users. Our department is committed to helping researchers find the most appropriate solutions to meet their research needs. The Beartooth cluster follows a condominium-style model of resource management. Condo computing resources are used simultaneously by multiple users. Researchers may work with ARCC to purchase compute nodes and storage which are then installed within the cluster and managed by our system administrators. When investor purchased resources are not in use by the prioritized investor users, they are made available as a community resource and general access to investor-purchased resources can be revoked whenever an investor wishes to use their resource. Prempted jobs are automatically re-queued to ease any burden on these general users. The Beartooth cluster currently offers over 500 compute nodes and over 1.2 Petabytes of storage. Our staff works to maintain an expected uptime of 98%, allowing for researchers to perform computation-intensive analyses on large data sets. The Beartooth compute environment may be accessed anywhere, anytime using SSH connectivity with UWYO two-factor authentication. In addition, the Wyoming Legislature has generously allocated the University of Wyoming with $5 million to expand our current hardware and infrastructure, and expanded capabilities are expected for full utilization by the middle of 2024.