এই এপিসোড সম্বন্ধে
CDFAM Computational Design Symposium — Washington DC 2026
Michael Frei · ARENA Physica
Most advances in computational design focus on mechanical structure — domains we can visualize and have evolved an intuition for. But as modern hardware becomes increasingly software defined, the unseen and unintuitive world of electromagnetism is taking center stage.
Conventional solvers can simulate fields, yet they cannot imagine new ones. Over the past year, our team has been building toward that capability. At CDFAM NYC and Barcelona, we shared early results from Atlas — an AI that learns electromagnetic behavior inductively from test data rather than deductively from first principles, enabling verification, optimization, and design postulation in domains where classical simulation reaches its limits.
This talk shares our vision for the future of AI-driven electromagnetic design.
Links
Talk page with full transcript
Search the full text of every recorded CDFAM presentation
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.designforam.com
