mauricio buendia

I'm fascinated by the physical substrate of intelligence: how electrons in silicon become computation, and eventually something that looks like thinking.

My master's at the Institute of Neuroinformatics (UZH / ETH Zurich) focuses on FeFET device modeling and FeFET-based recurrent neural networks in 22nm FD-SOI. Alongside it, I'm building Driffusion, a differentiable drift-diffusion simulator that uses gradients through semiconductor physics to inverse-design devices.

I've worked in space technology, and it's stayed with me. My curiosity runs from chips, radio, FPGAs and servers to supply chains and 24/7 industries, where everything comes back to real people and places.

I got here through music. A computers-and-music elective turned into a love of sound synthesis, then electronics, then everything that followed.