🚀 A Mars lander slamming on the brakes at Mach 2.4, volume-rendered live in a browser tab with CoDim. A simulation result like this usually lives trapped on a computer that costs more than a car. No screenshot. No desktop workstation. No massive file downloads. The actual exhaust plume, scrubbing frame-by-frame. What you're looking at: NASA's FUN3D simulation of supersonic retropropulsion. A 16.4 m heat-shield aeroshell fires 8 engines forward, straight into a Mach 2.4 supersonic wind. This is what we're making at CoDim. #scientificvisualization #CFD #Mars #aerospace #PyVista #3D #visualization
About us
At CoDimensional, we're building the missing collaboration layer for 3D data. Founded by the maintainers of the open-source tools your team already runs.
- Website
-
https://codimensional.com
External link for CoDimensional
- Industry
- Data Infrastructure and Analytics
- Company size
- 2-10 employees
- Type
- Privately Held
Employees at CoDimensional
Updates
-
Will this board cook itself? Here's heat-transfer simulation with solid components heating up and forced air carrying the heat away. Mechanical, electrical, and thermal analysts can stop trading three different screenshots of the same run and look at one live scene in #CoDim #ThermalManagement #CFD #ElectronicsCooling #HardwareEngineering #3D
-
This is Wendelstein 7-X, the largest stellarator ever built. The glowing surface is the plasma boundary, colored by magnetic field strength. No matter your domain, in this case nuclear fusion, CoDim can you help gain insight into your data. #fusion #fusionenergy #stellarator #scientificvisualization #computationalphysics #CoDim #PyVista
-
CoDimensional reposted this
So excited to see Laure Zanna's work on Samudra 2 visualized in CoDim. Samudra 2 is an AI ocean model that runs on a single GPU and is about two orders of magnitude faster than the traditional model it learns from. Read their write-up on what Samudra 2 is and why the scale matters: https://lnkd.in/gdn_GfgH Great work by the team at NYU, Open Athena, and collaborators. Stoked that we at CoDimensional were able to help make some stunning visuals. #OceanModeling #ClimateScience #ScientificVisualization #AI #3D #PyVista
-
CoDimensional reposted this
So excited to see Laure Zanna's work on Samudra 2 visualized in CoDim. Samudra 2 is an AI ocean model that runs on a single GPU and is about two orders of magnitude faster than the traditional model it learns from. Read their write-up on what Samudra 2 is and why the scale matters: https://lnkd.in/gdn_GfgH Great work by the team at NYU, Open Athena, and collaborators. Stoked that we at CoDimensional were able to help make some stunning visuals. #OceanModeling #ClimateScience #ScientificVisualization #AI #3D #PyVista
-
CoDimensional reposted this
We see this across so many sectors: AI is accelerating complex data production by 100x and more. The more data these methods generate, the more we need tools to inspect and analyze all of it together in teams. We loved helping with the video for this blogpost. If you want to explore data like this interactively, come talk to us.
Keller Professor of Applied Math @ New York University, Courant Institute + M²LInES Scientific Director
🌊 High-resolution ocean modeling on a single GPU with AI: meet Samudra 2 🧠 A neural network that imitates a world-class ocean model (GFDL's OM4) and runs ~100x faster at eddy-permitting resolution (1/4°)! ✅ Eddies and sharp currents like the Gulf Stream, that coarse models miss ✅ Cuts deep-ocean errors by 7–10× compared to Samudra at coarse resolution ✅ Tracks El Niño variability better than before 🌎 The real shift: when a multi-year run fits on one GPU, you stop rationing scenarios and start running thousands — the regime needed for ⚓ Shipping & offshore ⚡ Energy & insurance 🌾 Seasonal forecasting 💻 Fast, cheap, and high-resolution, fully open: code, and weights all available 👉 https://lnkd.in/eHBwnrm2 preprint: https://lnkd.in/eqgrpMWn #AI4Science #Climate #MachineLearning #OceanModeling M²LInES project M²LInES Schmidt Sciences Open Athena New York University NYU Courant Institute School of Mathematics, Computing, and Data Science NYU Center for Data Science YUAN YUAN Adam Subel Alistair Adcroft Carlos Fernandez-Granda Julius Busecke Alex Merose Jesse Rusak Pavel Perezhogin
-
We see this across so many sectors: AI is accelerating complex data production by 100x and more. The more data these methods generate, the more we need tools to inspect and analyze all of it together in teams. We loved helping with the video for this blogpost. If you want to explore data like this interactively, come talk to us.
🌊 High-resolution ocean modeling on a single GPU with AI: meet Samudra 2 🧠 A neural network that imitates a world-class ocean model (GFDL's OM4) and runs ~100x faster at eddy-permitting resolution (1/4°)! ✅ Eddies and sharp currents like the Gulf Stream, that coarse models miss ✅ Cuts deep-ocean errors by 7–10× compared to Samudra at coarse resolution ✅ Tracks El Niño variability better than before 🌎 The real shift: when a multi-year run fits on one GPU, you stop rationing scenarios and start running thousands — the regime needed for ⚓ Shipping & offshore ⚡ Energy & insurance 🌾 Seasonal forecasting 💻 Fast, cheap, and high-resolution, fully open: code, and weights all available 👉 https://lnkd.in/eHBwnrm2 preprint: https://lnkd.in/eqgrpMWn #AI4Science #Climate #MachineLearning #OceanModeling M²LInES project M²LInES Schmidt Sciences Open Athena New York University NYU Courant Institute School of Mathematics, Computing, and Data Science NYU Center for Data Science YUAN YUAN Adam Subel Alistair Adcroft Carlos Fernandez-Granda Julius Busecke Alex Merose Jesse Rusak Pavel Perezhogin
-
CoDimensional reposted this
You cannot screenshot a shockwave. This is a supernova blast wave: a hydrodynamics simulation rendered as log temperature so the shock-heated ejecta glow while the cold medium stays dark. In CoDim it plays back as a live volume in the browser, scrubbable frame by frame, interactive while it runs. The thing scientists actually care about here is time. A still throws that away. A rendered mp4 locks the camera and the colormap forever. A live scene lets the next person ask their own question of the same data. #ScientificVisualization #Astrophysics #ComputationalScience #3D #DataVisualization
-
You cannot screenshot a shockwave. This is a supernova blast wave: a hydrodynamics simulation rendered as log temperature so the shock-heated ejecta glow while the cold medium stays dark. In CoDim it plays back as a live volume in the browser, scrubbable frame by frame, interactive while it runs. The thing scientists actually care about here is time. A still throws that away. A rendered mp4 locks the camera and the colormap forever. A live scene lets the next person ask their own question of the same data. #ScientificVisualization #Astrophysics #ComputationalScience #3D #DataVisualization
