Over the last few years, the global high performance computing (HPC) community’s adoption of the Ohio Supercomputer Center’s (OSC) Open OnDemand platform has skyrocketed. With more research computing professionals discovering the benefits of offering a web browser portal to their HPC resources, usage has increased to more than 2,100 locations in 100 countries.
Texas A&M University’s High Performance Research Computing (HPRC) was an early adopter of Open OnDemand and has utilized the portal for its 3,000 faculty, staff, and student users. HPRC has extensively customized the platform, engages in global conversations with the HPC user community, and is a partner institution on a National Science Foundation (NSF)-funded project to help further develop Open OnDemand. The Texas A&M team has become a go-to resource for other HPC centers seeking to adopt or customize Open OnDemand, and it participates in activities such as the Global Open OnDemand (GOOD) conference to continue collaborating with others.
Today, Open OnDemand has been deployed on all five of Texas A&M’s clusters, which boast 12.5 petaflops in peak performance with 24.7 petabytes of high-performance storage. After discovering how easily Open OnDemand allowed university researchers to quickly access HPRC’s resources—with no software to install, a web browser interface, and a smooth connection to the clusters—HPRC made it the primary mode of access for its users.
Texas A&M has built a custom Open OnDemand dashboard for its community, which provides a quick overview of the number of jobs running, as well as core, node, and disk quota utilizations. Users can quickly file and track help tickets and review available software within the environment before requesting support. The latest version of the dashboard now allows users to drag and drop elements to design a custom interface.
One of Texas A&M’s most notable contributions to the Open OnDemand platform is its development of a suite of apps that help research computing professionals manage resources, such as those that integrate help desk ticket software, resource allocation dashboards, and job schedulers. The HPRC team also has created apps to support its research community, such as a custom version of CryoSPARC, which supports cryo-electron microscopy (cryo-EM) used in research and drug discovery work, and an interactive app for the health sciences faculty. The team also has worked on the Jupyter AI Assistant, which supports coding on Jupyter Notebooks.
Another example is the Drona workflow engine, which allows researchers to design custom workflows in research computing, especially for running AI workloads. In addition, HPRC developed a status app to help users find available GPU nodes, which reduces wait times and improves utilization of the research computing clusters. Marinus Pennings, associate director for user support and helpdesk, and Michael Dickens, research scientist, of the Texas A&M HPRC team presented these innovations at the 2025 GOOD conference at Harvard University.
Dhruva Chakravorty, director for research and emerging technologies, is now leading an effort to make apps available for scientific applications and research workflows for the broader Open OnDemand community. As co-principal investigator on the NSF-funded Growing Open OnDemand: Leveraging Unified Community Knowledge (GOODLUCK) project, Chakravorty is overseeing the curation of a suite of apps on the Open OnDemand site called the Appverse. The Appverse features dozens of software apps with links to repositories with code and setup details so that HPC centers may view the various ways institutions have implemented them for their users.
Chakravorty and several other members of the Texas A&M HPRC team also regularly participate in national conversations with the Open OnDemand community, such as by presenting about their dashboard and apps at the GOOD conference in March 2025.
“People are invested in helping each other,” Chakravorty said. “It’s a huge community-driven effort.”
Lisa Perez, director for advanced computing enablement, also appreciates the strong communication and easy exchange of information between the members of the Open OnDemand community, as well as the OSC staff.
“I only had to reach out for support on a couple of items, and the response from the admins has always been very helpful,” Perez said.
Looking forward, the Texas A&M HPRC team is setting its sights on tackling several major issues in research computing, with the support of Open OnDemand. The staff is investigating how to best support artificial intelligence (AI) and quantum computing workflows, as well as cloud integration, in their operation. Inspired by OSC, the Texas A&M team also would like to deploy Open OnDemand on Kubernetes for their classroom users.
The HPRC staff also will be charged with implementing Texas A&M’s major new investment in supercomputing, a $45 million NVIDIA DGX SuperPOD that is expected to be one of the top-performing academic systems in North America. With 760 NVIDIA H200 GPUs scaled with NVIDIA Quantum-2 InfiniBand networking, the new system will support generative AI, machine learning, and scientific simulations.
“HPRC is leading the charge to support the growing demand for AI and machine learning research and education tools at Texas A&M,” said Honggao Liu, executive director of the HPRC. “We anticipate that Open OnDemand will play a key role in our efforts.”
Written by Andrea Gibson
The Ohio Supercomputer Center (OSC) addresses the rising computational demands of academic and industrial research communities by providing a robust shared infrastructure and proven expertise in advanced modeling, simulation and analysis. OSC empowers scientists with the services essential to making extraordinary discoveries and innovations, partners with businesses and industry to leverage computational science as a competitive force in the global knowledge economy and leads efforts to equip the workforce with the key technology skills required for 21st century jobs.
