High performance computing for advanced manufacturingΒΆ

For the last two years I have been exploring how to leverage high performance computing capabilities for advanced manufacturing applications.

Through my Madman project (Madman=models for advanced manufacturing), I have been developing models in the following three areas:

  • Predictive scale up of atomic layer deposition, such as the spatial ALD on high surface area materials.
  • Optical and electronic materials and devices, from coatings for concentrated solar power to wide bandgap semiconductors.
  • Physics-based models for additive manufacturing of metals.

The focus here is not only to take advantage of many cores to carry out complex simulations, but also to run hundreds or thousands of independent simulations in parallel, for instance to efficiently explore a parameter space or to generate large datasets that can be then analyzed using deep learning approaches.