PARCC’s Betty Cluster Goes Live

Hello Wharton Researchers!

The Penn Advanced Research Computing Center (PARCC)’s Betty Cluster went online in production beginning on Sunday, March 15th. It will become a central resource for research computing at Penn. Wharton is actively partnering in its rollout and will be gradually transitioning our research workloads there from our current HPC3 environment. Wharton researchers will receive:

  • Continuing hands-on support, onboarding assistance, and training from the Wharton Research & Analytics team.

  • Subsidized access to substantial compute and storage resources.

  • Options to expand via paid capacity.

Read on for more details, and please look for further direct contact in the next few weeks from our team to discuss your team’s needs and participation. You are of course welcome to reach out to us at research-computing@wharton.upenn.edu, as well.

About PARCC and the Betty Cluster

PARCC is a new center at Penn, intended to serve as a centralized computational resource available to all twelve schools at Penn. This will include support for system administration and data science consulting services that cater to our diverse research computation needs.

The Betty Cluster offers a large set of compute resources, storage, and research software for your use.

PARCC’s Betty cluster rates are set so that it can perpetually fund improvements, upgrades, maintenance, operations, and staffing. Over time, PARCC will continue to grow and keep pace with the latest in cluster technologies.

More details on the services available can be found at Penn Advanced Research Computing Center (PARCC).

Wharton Participation in PARCC

We are working closely with the PARCC Team to increase the school’s use of PARCC resources over time, eventually transitioning our current cluster workloads to Betty.

Subsidized Resources for Wharton Researchers

Wharton Computing’s Research & Analytics Team (my team!) will be working with each interested Wharton researcher and their team to gain access and begin doing research using PARCC Betty cluster resources.

As with our current and past Wharton cluster, we will be covering the cost of a generous allocation of Betty cluster resources for your research needs, including 50GiB of storage per user, 1TiB of storage per PI / Center (Lab / Project space), and a similar or greater-sized set of CPU and GPU resources than we currently provide for you.

Any software that’s not available to the general Betty population that is needed by our researchers will be considered for installation and maintenance by Wharton Research and Analytics for your use.

Wharton datasets that are currently available on the HPC3 will also be available on Betty.

Pay-As-You-Go Resources for Wharton Researchers

Also similar to our current cluster (HPC3), faculty PIs and Centers will be able to purchase resources beyond the fully-subsidized set of resources available. This is how other schools’ researchers will generally be using PARCC resources.

Transition from Wharton’s HPC3 and Onboarding for Wharton Researchers

  • We will be reaching out to all current Wharton HPC3 researchers over the next few weeks to discuss your ongoing cluster needs.
  • We and the PARCC team will work with our researchers and their teams to transition or onboard to PARCC as smoothly as possible, taking into account currently-running work, and personnel schedules.
  • We and the PARCC team will be coordinating offerings of Q&A sessions, training, and documentation about the new systems.
  • We will continue to maintain the current Wharton cluster (HPC3) until it’s no longer needed, and the PARCC transition is considered successful and complete.

Ongoing PARCC and Research Support

We are and will continue to work very closely with the PARCC team to make sure that our researchers are getting the most out of this resource.

We will be an integral part of the PARCC support model, able to assist you and your team with requests and the ability to participate in their ticketing system and attend regular meetings with the PARCC administrative team.

Our team will be using the resources ourselves, both in support of our researchers’ work, and for administrative purposes.

Q&A / FAQ

We have begun building a PARCC Q&A Page with answers to some of the most-commonly asked questions that we have received, and will be adding to that over time.

Please don’t hesitate to reach out to research-computing@wharton.upenn.edu to begin onboarding, or with questions not answered here.

About

With three decades of experience supporting research and more than two decades at The Wharton School, Hugh enjoys the challenges and rewards of working with world-class researchers doing Amazing Things with research computing. Robust and scalable computational solutions (both on premise and in The Cloud), custom research programming solutions (clever ideas, simple code), and holistic, results-focused approaches to projects are the places where Hugh lives these days. On weekends you're likely to find him running through the woods with a topo map and compass, orienteering, or participating in ultramarathon trail running events.