CHICAGO, IL // November 15-20, 2026

Kitware at Supercomputing 2026

At SC26, Kitware will showcase interactive visual workflows designed for large, complex simulations. Our work combines scalable visualization, in situ analysis, AI, GPU computing, and modern application development to support today’s demanding scientific and engineering workflows.

Visit Kitware’s booth to see these capabilities in action and talk directly with our technical team.

// Details

McCormick Place: Booth 5226

Presentation Schedule

Presentation

The State of Open Source Computer Architecture Simulators: Shared Problems and Community Action

Nov 19 | 12:15PM-1:15PM CT W183a

Caitilin Ross

Computer architecture simulators are central to exploring the design space of modern high-performance hardware and analyzing its performance. Many open source simulators exist, targeting different hardware scopes, granularities, and metrics, yet none is perfect: issues with speed, accuracy, and usability limit their effectiveness for architecture researchers, accelerator designers, hardware-software co-designers, and performance engineers, among others. This BoF convenes an expert panel of users and developers from academia and industry to discuss these tools’ shortcomings and possible responses. We invite all who build, use, or depend on these simulators to join the discussion and identify impactful community efforts, with outcomes shared publicly.

A web browser interface for Model-assisted annotation built with trame, featuring a 2D slice segmentation view alongside a 3D volumetric rendering of multi-class material segmentation (CFO, CGO, Gd-rich emergent) with AI active learning controls.

Demo

AI for Science with tomviz-web

Nov 17-18 | 10am-6pm CT and Nov 19 | 10am-3pm CT Booth 5226

Ask an AI agent to explore 3D microscopy data and watch it drive tomviz-web, an interactive web application. Then flip the loop: mark up a few images in tomviz-web, and an AI model applies that guidance across the whole dataset.

A 3D computational fluid dynamics visualization rendered in ParaView showing helical flow streamlines extending from a blue nozzle orifice, color-coded by velocity or scalar values ranging from white to pink and dark red

Demo

GPU-Enabled Visualization

Nov 17-18 | 10am-6pm CT and Nov 19 | 10am-3pm CT Booth 5226

See how GPU computing can support more robust, highly interactive visualization and analysis for demanding scientific datasets.

A complex 3D aerodynamic simulation in ParaView depicting turbulent flow and vortex shedding around a multi-blade rotor or fan assembly, featuring volumetric velocity fields and slice plane contours in blue, copper, and white.

Demo

Catalyst and In Situ Workflows

Nov 17-18 | 10AM-6PM CT and Nov 19 | 10AM-3PM CT Booth 5226

Explore approaches for bringing analysis closer to running simulations to reduce data movement and post-processing bottlenecks.

A two-panel scientific visualization figure showing a curved stepped grid with swirling flow streamlines on the left, and a porous blue material matrix with yellow and red tracer particles on the right.

Scalable Simulation Analysis

As simulations grow in size and complexity, visualization and analysis must scale with them. Kitware builds scalable pipelines for large datasets, complex results, and demanding computational environments. We design these workflows to maintain interactivity and analytical depth as workloads increase, so scientists and engineers can continue to explore, validate, and understand simulation results efficiently.

A two-panel scientific visualization showing a subsurface wellbore trajectory and microseismic point cluster on the left, and a global earth globe with a mapped equatorial atmospheric or oceanic grid overlay and UI controls on the right

Data Movement Optimization

Moving large volumes of simulation data can put significant pressure on storage, I/O, and downstream analysis. Kitware redesigns data workflows to reduce unnecessary movement and make better use of available computing resources. We look at how data is generated, moved, stored, and analyzed to create more efficient workflows from simulation through post-processing.

A two-panel aerodynamic CFD visualization displaying full-aircraft wake streamlines and surface pressure distribution on the left, alongside 2D/3D laminar airflow streamlines passing over an isolated airfoil section on the right.

Simulation-to-Decision Workflows

Modern HPC workflows depend on more than simulation performance alone. Kitware connects computation with analysis, AI, and decision-making so results can move more efficiently through the full workflow. By reducing bottlenecks between these stages, we help shorten the path from simulation output to useful scientific and engineering insight.

A two-panel scientific visualization figure showing helical flow streamlines exiting a blue nozzle on the left, and volumetric vortex shedding around a multi-blade fan assembly on the right.

In Situ Analysis

ParaView Catalyst brings analysis into the simulation workflow so data can be examined and processed as it is generated. Kitware integrates Catalyst into existing simulation environments and identifies where in situ analysis can provide the most value. This can reduce reliance on traditional post-processing and support scalable analysis for demanding HPC workloads.

A two-panel scientific visualization figure displaying a matrix grid of global climate model maps on the left, and a 2D turbomachinery cascade simulation showing fluid flow across curved turbine blades on the right.

Interactive Visual Workflows

Computation, visualization, and interaction are often handled as separate stages, which can make scientific workflows harder to manage. Kitware designs visual workflows that bring these capabilities together around the needs of the application. This gives teams a more connected way to explore data, understand results, and carry out analysis.

A two-panel scientific visualization figure showing a high-detail CAD model of a Star Wars TIE Advanced starfighter on the left, and a 3D geological elevation model with stratified subsurface soil layers on the right.

Web-Based Simulation Monitoring

Trame and Catalyst make it possible to monitor simulation results through a web browser using remote rendering. Kitware builds web-based workflows that provide access to live visualization across devices without requiring specialized client software or a powerful GPU on every machine. This makes simulation monitoring more accessible while keeping the computational work close to the resources where it is running.

Related Projects

A 2D computational fluid dynamics simulation showing streamline flow and scalar velocity contours over a backward-facing step or curved channel cavity, featuring a prominent recirculation vortex in the lower blue region.

NASA: Bringing AI Into Aerospace Simulations

Kitware is building open source AI tools that plug into NASA’s existing aerospace simulations, helping engineers run studies faster, explore more scenarios, and reduce computing costs.

A 3D scientific visualization illustrating an overset grid or digital twin marine CFD workflow, showing a boat navigating surface waves on the left alongside an isolated hull mesh inside a transparent bounding subdomain box on the right.

Wave Tank Simulation Workflows with CMB and ParaView

CMB provides a task-based workflow system for OpenFOAM simulations, streamlining the process from model setup to result analysis while reducing errors and wasted HPC cycles.

A 3D geospatial scientific visualization displaying an orthographic globe centered on the Americas, wrapped with a dense, color-mapped equatorial structured grid alongside an interactive control panel on the left.

Pan3D: VTK-Powered Interactive Climate Analytics

Pan3D brings interactive analysis and 3D visualization into the same workflow, helping researchers explore large geophysical datasets without relying on custom data conversions or disconnected tools.

$A 3D CONVERGE CFD combustion simulation inside an internal combustion engine cylinder showing formaldehyde mass fraction isosurfaces in yellow, red, and blue beneath intake valves.$

Reducing Simulation I/O with In Situ Analysis

In situ analysis brings visualization and analysis into the simulation itself, reducing file writing bottlenecks and making results available sooner. In one case, it cut simulation run time by 20% and reduced output from 100GB to around 100MB.

A 3D schematic model in ParaView depicting battery microstructure layers labeled Separator, Electrode (with cylindrical active material particles in dark red), and Current Collector.

Battery Simulation Workflows with trame

Using trame, we transformed a manual battery simulation into a streamlined web application. The result is an easier way to configure, run, and analyze advanced simulations.

A 3D subsurface reservoir visualization enclosed in a wireframe bounding box, showing curved wellbore trajectories intersecting a dense cluster of microseismic event points against a background point grid colored red and blue.

EarthSight: Subsurface Visualization with ParaView

EarthSight uses ParaView and VTK to create an intuitive application for subsurface modeling and visualization. The workflow simplifies complex geoscience data while preserving advanced 2D and 3D visualization capabilities.