JuliaHub's fight for engineering workbench

The real fight is over who gets to be the engineer's default workbench before any single model is built. Inside big industrial accounts, incumbent vendors already own the budget, the approval path, and years of validated models. On the edge of that stack, cloud and HPC vendors can still capture new spending by hosting existing tools faster and cheaper, which means JuliaHub has to win both a workflow migration and an infrastructure wedge at the same time.

• MathWorks is hard to displace because MATLAB and Simulink are already tied into model libraries, code generation, verification, and engineer training. MathWorks has also kept moving up the stack with MATLAB Online, AWS cloud cluster management, and Simulink Copilot, so JuliaHub is not attacking a stagnant incumbent.
  1 sacra 2 mathworks 5 mathworks
• The industrial giants bundle simulation into much larger software estates. Siemens closed Altair in March 2025, adding HPC and simulation to Xcelerator, and Synopsys closed Ansys in July 2025, deepening its digital twin and multiphysics reach. In those accounts, buying decisions often follow the existing PLM and engineering software standard.
  1 sacra 3 mathworks 6 siemens 7 synopsys
• Rescale shows the infrastructure threat clearly. It offers a cloud layer with thousands of pre installed engineering packages, so an enterprise can modernize how jobs run without changing modeling tools. JuliaHub's answer is Dyad and migration services, which try to turn infrastructure entry into a full modeling platform switch.
  1 sacra 4 rescale

Going forward, the strongest openings are programs where speed, cloud deployment, and AI assisted model building matter more than staying inside one incumbent stack. If JuliaHub keeps landing migration projects and embeds alongside tools like Ansys rather than waiting to replace them outright, it can expand from a specialist layer into a broader engineering platform.