The day in conversation for 2026-11-08

discussion

Transcript

The discussion: Kai: The acceleration claimed for SesQ, going two orders of magnitude over FEM by reducing dimensionality to quasi-2D layers, seems huge, but I’m worried about how that translates when you move from idealized qubit layouts to real fabrication constraints where edge effects become dominant. Mira: That dimensionality reduction is impressive for the simulator, but what about the practical fidelity of those surface integral equations when dealing with non-uniform material properties inherent in superconducting circuits? Are we assuming a uniform substrate dielectric constant across the entire simulated layer? Lev: If we assume uniformity, then the computational speedup is robust for that specific problem space, but I wonder if that focus on minimizing design-dependent losses misses potential bottlenecks when considering how those simulated states interact with the classical nonlinear dynamics modeled in the Kicked Top analysis. Kai: Exactly. The Kicked Top model explores classical chaos, and if SesQ is primarily solving static loss minimization,

More episodes

← Home