News, tutorials, and deep dives from the Tesseract team.
A training loop can now call a GPU solver running in another process, forward and backward, without its arrays ever leaving the GPU. Here is what that takes, what it buys, and how it works.
Announcing the winners and highlights from the 2026 Tesseract Hackathon.
Coupled multi-physics used to mean a hand-written adjoint or a monolithic rewrite. Compose independent Tesseracts and jax.grad differentiates the whole two-way coupling end-to-end.
Introducing cookiecutter-tesseract, a batteries-included starting point for building projects out of more than one Tesseract.
How we built a pipeline to get exact gradients out of a Fortran thermal solver, then used it to solve real inverse problems from Python.
We show how to chain adjoint, finite-difference, and AD gradients across Ansys SpaceClaim, PyMAPDL, and JAX to optimize rocket grid fins end-to-end.
The inaugural Tesseract Hackathon concluded with impressive entries from global researchers and engineers.
A hands-on tutorial introducing the Tesseract ecosystem through a physics-based example.
A demo highlighting how Tesseracts unlock efficient, clean pipeline-level automatic differentiation.
We're announcing the public release of Tesseract Core, a free and open source application enabling scientists and engineers to build end-to-end differentiable pipelines with minimal code.