Researcher Forum 2026 Highlights

Read on for highlights from the 2026 CHTC Researcher Forum, summarizing last year’s activities and plans for the upcoming year.

Table of Contents

CHTC Community Updates

Community knowledge sharing

We want to help CHTC users share their accumulated knowledge with each other - whether it’s general CHTC tips or experience specific to your domain.

🧬 Bioinformatics Café

This community-of-practice started in Fall 2025 and brings together life-science researchers using CHTC resources. Monthly meetings include tutorials and facilitation support. Beginners are welcome!

For more information, see this page: Bioinformatics Cafe.

✏️ CHTC Community Forum

(Not to be confused with the Researcher Forum event!)

The facilitation team launched a community forum platform to facilitate communication between CHTC users. Any CHTC user can sign in (no account creation needed) and ask questions, share ideas, and search previous posts from other users to learn more about using CHTC to power your research.

The facilitators also use the forum to post announcements about upcoming events and opportunities.

To get started, login to community.chtc.wisc.edu using your NetID.

Integrating data with computing

Research computing is at its best when empowered by real-world data; with the Pelican Platform (developed at CHTC!), CHTC hopes to streamline the process of getting data to computing.

💾 CHTC and ResearchDrive

With the new “UW Data Federation (UWDF)” integration with ResearchDrive, powered by the Pelican Platform, users can now declare in their HTCondor submit files how each job should transfer data to/from ResearchDrive. Input data in ResearchDrive is transferred automatically at the start of the job, while output data can be transferred back to ResearchDrive at the end of the job.

To get started, read the guide. We’re also hosting a dedicated training on April 7th.

Do you have a large volume of data not in ResearchDrive, and want a similar functionality for your HTCondor jobs? Let us know - we may be able to help!

AI at CHTC

By leveraging the power of High Throughput Computing (HTC), we’re working to make CHTC a key part of the campus ecosystem for AI support.

HTC and AI

HTC works best at executing many independent calculations, for example: inference workflows, (hyper)parameter tuning, and dataset preparation. While not well supported at this time, we are exploring how we can help users interested in interactive code development and testing.

Facilitation and AI

The facilitation team has been developing tutorials and guides for users interested in using CHTC for AI research computing.

In addition to our general facilitation support, we also have an expert facilitator who is willing to consult on AI workflows.

GPUs at CHTC

Thanks to university funding, CHTC regularly purchases GPU servers to increase our open capacity. For a list of the GPUs available in CHTC account, see the table here. These GPUs are available to anyone with a CHTC account, with the option to scale beyond the open capacity as described here.

Access compute from a browser

We’re working to make it possible to access powerful computing from a web browser.

🖥️ HPC and Open OnDemand

Users of the High Performance Computing (HPC) “Spark” cluster can now access the system via Open OnDemand (openondemand.org). Let us know if there are any features you would like to see beyond this base deployment.

To get started, read the guide: Open OnDemand.

💻 BadgerCompute

In partnership with DoIT and the Data Science Institute, we’ve launched a Jupyter-based interactive computing service for researchers on campus. Compute, visualize data, and write Jupyter notebooks without additional installation or leaving your laptop.

For more information on the service, see badgercompute.wisc.edu.

To get started, read the guide.

Improvements to HTCondor

CHTC is the home to HTCondor, a powerful scheduling tool used not only at CHTC but at 100s of organizations around the world. As HTCondor users at CHTC, you have a unique opportunity to share your ideas with the development team (via the facilitators) to drive improvements and changes. Here are some of the recent or proposed changes motivated by user and facilitator feedback.

Deployed

These features have been released and are ready to use.

  • write your job command into the submit file - with the new shell option, users can now declare the command they want their job to execute directly in the submit file. See the CHTC guide or HTCondor manual.

  • more options for retrying jobs with varying memory needs - users can now provide a second, larger memory request (retry_request_memory) for jobs that exceed the initial memory request, and HTCondor will automatically retry the job with the larger amount. See the CHTC guide or HTCondor manual.

  • improvements to condor_watch_q (a user favorite!) - file transfers are now included in the output; easier to exit using any key.

Active development

These features are being worked on and may have beta versions available for testing. Contact the facilitation team if you are interested in testing.

  • “common files” transfer - users can declare a set of files that every job needs and HTCondor will automatically reuse said files across multiple jobs on the same machine.

  • tool for resource usage analysis - users can see a summary report of the amount of resources their jobs actually used, to help optimize the amounts that they are requesting.

  • submit jobs from your computer - using “remote submission,” users can submit jobs directly from their own computer (instead of an Access Point) using Python.

In design phase

These items are under consideration for future development. Contact the facilitation team if you are interested in participating in the conversation.

  • “noun verb” commands interface - interact with HTCondor using an easier-to-learn set of commands that are more consistent with other CLI tools.

  • finer-grained declaration of file transfers - declare multiple file transfers and mappings more easily, while preserving backwards compatibility.

  • more support for interactive work - deploy interactive computing environments more easily.

Disclaimer: these items are not guaranteed to be implemented.

Facilitation improvements

Now at four full-time staff, the facilitation team continues to improve user support resources.

Website updates

In the past year, we’ve:

  • revamped the website navigation and standardized guide formats
  • streamlined guides on using containers
  • added feature-specific guides

along with a variety of other small changes.

Looking forward, we’re planning to add a “hello world” guide on how to build containers and to revamp the landing page for user guides.

Check out the CHTC Research Computing website.

CHTC Recipes

We provide a central GitHub repository with examples of software containers and less common workflows, at github.com/CHTC/recipes. In this past year, we added several more recipes, updated several others, and had our first user contributions.

Looking forward, we plan on adding more recipes and hope to pull in more user contributions!

Bioinformatics support

In addition to the Bioinformatics Café discussed above, we’re working to support bioinformatics researchers on campus and elsewhere.

The databases necessary for running AlphaFold3 have been populated on many of our HTC system servers; users no longer have to worry about managing the 700+ GB of data needed for their AlphaFold jobs. We’re exploring how to make these databases available as part of the OSPool and considering what other bioinformatics databases we can host for users.

Bioinformatics research increasingly uses workflow managers Nextflow and Snakemake for executing data processing pipelines. We’re investigating how we can support these programs on CHTC so that researchers can continue to use tools that they are familiar with.

Save the dates - Summer 2026

CHTC hosts two (inter)national events in support of research computing each summer.

HTC26

This year’s Throughput Computing Week (HTC26) is being held June 9 - June 12 in Madison. Throughput Computing Week brings together researchers, campuses, international science collaborations, and more in a multi-day conference. Learn what’s new in HTC, how others employ the technologies to power their research, and connect with research computing professionals from diverse communities.

For more information, see the announcement here.

OSG School 2026

The OSG School is a week-long summer school where researchers (and research staff) learn how HTC systems work and how they can apply tools like HTCondor and the OSPool to their own research or campus.

This year, the School is July 13 - July 17.

Applications are open between February 20 and March 20, 2026. Visit osg-htc.org/school-2026 to learn more.

Keynote speaker: Ariana Negreiro

A recent graduate of UW-Madison, Dr. Ariana Negreiro used CHTC extensively to conduct the data processing and analysis that was central to her PhD dissertation. Dr. Negreiro worked closely with CHTC staff to develop her workflow and to set up the data transfer pipeline.

For more information on Dr. Negreiro’s research, read this article.

About CHTC

The Center for High Throughput Computing (CHTC) is more than just a compute cluster - we’re a research center dedicated to empowering researchers worldwide through the paradigm of High Throughput Computing (HTC).

For more information, see our main website at chtc.cs.wisc.edu.

Share your feedback

We want to hear from you!

We are strong adherents to the philosophy of “translational computer science”, which means that a lot of our work is based on user feedback. We are particularly interested in addressing “pain points” - things that are difficult, things that don’t work the way you expect, and things that you wish were easier. (We’re also happy to celebrate your successes for things that work well!)

Please feel free to reach out to the facilitation team with your feedback!