A many-core capable lightweight in-situ visualization and analysis infrastructure for multi-physics HPC simulations.
Ascent is a system designed to meet the in-situ visualization and analysis needs of simulation code teams running multi-physics calculations on many-core HPC architectures. It provides rendering runtimes that can leverage many-core CPUs and GPUs to render images of simulation meshes.
Ascent focuses on ease of use and reduced integration burden for simulation code teams.
- It does not require GUI or system-graphics libraries.
- It includes integration examples that demonstrate how to use Ascent inside three HPC-simulation proxy applications.
- It provides a built-in web server that supports streaming rendered images directly to a web browser.
Ascent Project Resources¶
Website and Online Documentation
Githib Source Repo
- Cyrus Harrison (LLNL)
- Matt Larsen (LLNL)
- Eric Brugger (LLNL)
- Jim Eliot (AWE)
- Kevin Griffin (LLNL)
- Hank Childs (LBL and UO)
- Utkarsh Ayachit (Kitware, Inc)
- Sudhanshu Sane (UO)
- Quick Start
- Ascent User Documentation
- Ascent Overview
- Example Integrations
- Building Ascent
- Build Dependencies
- Getting Started
- Build Options
- Host Config Files
- Building Ascent and Third Party Dependencies
- Uberenv Options for Building Third Party Dependencies
- Compiler Settings for Third Party Dependencies
- Building with Uberenv on Known HPC Platforms
- Building Third Party Dependencies for Development
- Building with Spack
- Using Ascent in Another Project
- Building Ascent in a Docker Container
- Building Ascent Dependencies Manually
- Asking Ascent how its configured
- Ascent API
- Ascent Actions
- Ascent Actions Overview
- Ascent Utilities
- Ascent GPU Notes
- Developer Documentation
- Ascent Contribution Guide
- Ascent Developement Overview
- Setting Up A Development Environment
- VTK-h Filter Anatomy
- Flow Filter Anatomy
- Publications and Presentations
- License Info