Visualization
Research and development of scalable, intelligent visualization and analytics for complex data.
Oak Ridge National Laboratory’s (ORNL’s) Visualization Group advances scientific discovery through innovative visualization and uncertainty quantification techniques for complex data generated by the world’s most powerful supercomputers. The group’s research combines cutting-edge approaches in ML, high-performance computing (HPC), and interactive visualization to help scientists better understand and communicate their results while ensuring the reliability and efficiency of next-generation computing systems. Through this work, the group aims to accelerate scientific breakthroughs across disciplines—including physics, biology, and materials science—while making supercomputing more accessible and sustainable.
The group’s overall mission focuses on several key areas:
- Visualization and uncertainty quantification
- Use of advanced computing (HPC, ML, quantum)
- Emphasis on scientific applications
- Making complex data more understandable
- Cross-disciplinary impact
- Connection to supercomputing facilities at ORNL
The Visualization Group is advancing scientific visualization and HPC to tackle some of today’s most complex scientific challenges. The group’s work spans three main areas: analysis and visualization of scientific data, visualization algorithms for next-generation hardware, and scientific workflow optimization.
The group’s research is pushing the boundaries of both quantum computing and traditional supercomputing: the group has developed predictive tools for quantum computing applications and is actively working on digital twin methods for ORNL’s Frontier machine—one of the world’s fastest supercomputers. Additionally, the group’s Viskores software library enables scientific visualization on exascale computers to help researchers more effectively leverage these powerful HPC systems.
The group has also conducted fundamental research on using visualization as a tool to help scientists better understand and visualize uncertainty in their data—a critical need as scientific simulations become increasingly complex. The uncertainty-aware techniques that the group created help scientists make informed decisions during their analysis.
Their work also includes novel techniques for analyzing critical points in scalar fields and managing uncertainty in neural networks to make complex data more interpretable and reliable. The breadth of impact goes well beyond the group’s own applications, and their underlying work supports advances in fusion energy research, materials science, and biological studies.
Looking ahead, the group will play a vital role in shaping the future of scientific computing and visualization. Their integrated approach—combining cutting-edge algorithms, ML, and interactive visualization—works toward a future in which complex scientific data becomes more accessible and actionable, accelerating the pace of scientific discovery across multiple domains.
