TACCSTER 2026

September 21-22, 2026

Texas Advanced Computing Center

Pickle Research Campus | Austin, TX


Speakers

Dr. Stefan Henneking

Oden Institute for Computational Engineering and Sciences
The University of Texas at Arlington

Forecasting Tsunamis: Real-Time Inference and Optimal Sensor Placement at Extreme Scale

Dr. Stefan Henneking is a Research Scientist at the Oden Institute for Computational Engineering and Sciences at The University of Texas at Austin. His research focuses on high-performance computing, scalable algorithms, and Bayesian inverse methods for complex scientific applications, including acoustic sensing, fiber optic modeling, and tsunami early warning. He is a recipient of the 2025 ACM Gordon Bell Prize.

Dr. Jeanne Kowalski-Muegge

The University of Texas Dell Medical School

Supercomputing Precision Medicine: Purpose-Built Infrastructure for Clinical AI

Jeanne Kowalski-Muegge, Ph.D. is Associate Director, Clinical AI for Precision Medicine and Professor, Department of Internal Medicine, Dell Medical School at The University of Texas at Austin. A leader in clinical AI and clinical cancer genomics, she advances their intersection to inform differential diagnosis and treatment strategies for complex care.

Dr. Jose Rizo-Rey

The University of Texas Southwestern Medical Center

The local detergent model of biological membrane fusion

Josep Rizo (Jose Rizo-Rey), Ph.D. is the Virginia Lazenby O'Hara Chair in Biochemistry and Professor of Biophysics at UT Southwestern Medical Center. His research uses structural biology and biophysical techniques to study the molecular mechanisms of neurotransmitter release and intracellular membrane fusion, advancing our understanding of how nerve cells communicate.

Dr. Dan Stanzione

Associate Vice President For Research, The University of Texas at Austin
Executive Director, Texas Advanced Computing Center

State of the Center Address

Dr. Dan Stanzione is a nationally recognized leader in the field of high performance computing who has made an impact in the open science community for more than 30 years.

He is Executive Director of the Texas Advanced Computing Center (TACC), one of the leading advanced academic computing centers in the world and holds the position of Associate Vice President for Research at The University of Texas at Austin.

Stanzione serves as the principal investigator (PI) for the U.S. National Science Foundation (NSF) Leadership Class Computing Facility (LCCF), which started construction in July 2024 and is expected to begin operations in 2026. The LCCF will deploy Horizon, which will be the largest academic supercomputer dedicated to open science research in the NSF portfolio.

Stanzione is also the PI for several current NSF-funded supercomputers including Frontera, the fastest supercomputer at a U.S. university, and Vista, an AI-centric system—all dedicated to open science research.

Stanzione received his bachelor’s degree in electrical engineering and his master's degree and doctorate in computer engineering from Clemson University.

Dr. Thomas Yankeelov

The University of Texas at Austin

High-performance computing to enable patient-specific digital twins in oncology

Tom Yankeelov is the W.A. “Tex” Moncrief Chair of Computational Oncology and Professor at The University of Texas at Austin. He directs the Center for Computational Oncology, where his research combines advanced medical imaging, patient-specific data, and computational modeling to improve cancer diagnosis, treatment, and patient outcomes.

Back to TACCSTER Overview

Tutorials

Building Interactive Scientific Dashboards in the Cloud with Plotly Dash

Presenter: Erik Ferlanti and James Labyer

In this workshop, we will dive into Plotly Dash, a powerful framework for building interactive web applications and dashboards in Python. We will learn how to create and customize Dash applications, and how to use them to visualize and explore scientific data in real-time. We will also learn how to deploy a Dash app into production on the Jetstream2 cloud.

Introduction to Horizon and CPU–GPU Architecture

Presenter: Dr. Lars Koesterke

The tutorial focuses on transformer-based neural operators, with Transolver and GeoTransolver as the worked examples. We will first build intuition for why transformers are useful beyond language: attention provides a way for distant points in a mesh or point cloud to exchange information, while physics-aware attention compresses large domains into a small set of learned physical states, making transformer models practical for large scientific datasets. We then connect these ideas to an end-to-end workflow using NVIDIA PhysicsNeMo on GPU/HPC systems: preparing simulation data, training, and validating a surrogate, interpreting accuracy, and using the trained model for rapid inference and downstream design studies.

This tutorial introduces Horizon, TACC’s latest high-performance computing system for open science, and uses it as a starting point for a broader discussion of modern HPC architecture.

If time permits, participants will have an opportunity to log in and explore a few short, practical exercises. For users who have not yet worked on Vista, this provides a useful bridge to Horizon, as the software environment, batch system, and basic workflow patterns are very similar.

Surrogate Modeling on TACC Systems

Presenter: Dr. Luke Smith

Surrogate modeling is a popular framework in which high-resolution simulation data is leveraged to build lean, inexpensive models for engineering and design workflows. This tutorial will guide attendees through the process of building a surrogate model from the output of a numerical simulation, with an emphasis on implementation strategies for TACC systems. As a driving example, we consider a dataset drawn from epidemiology, in which disease spread is predicted over time, and discuss options for storing data, leveraging existing neural network architectures, and making optimal use of TACC's GPU hardware. This tutorial is intended as a complement to ''Transformer-Based AI Surrogates for Scientific Simulations,'' with the aim of exploring additional resources available to TACC users.

Transformer-Based AI Surrogates for Scientific Simulations

Presenter: Dr. Karthik Mukundakrishnan

Scientific and engineering workflows often require hundreds or thousands of simulations for design exploration, uncertainty quantification, and optimization, but high-fidelity solvers can make that scale impractical. This hands-on tutorial introduces AI surrogate modeling as a practical way to learn fast approximations of simulation solution operators: mappings from geometry, boundary conditions, and operating parameters to full physical fields.

Back to TACCSTER Overview