Stan Tomlinson was a bright new PhD out of UCSD CogSci (?) who, with others. created a startup to pursue his thesis ideas about digital neural network architecture, DNNA. This involved doing multiply and accumulate stochastically with AND gates and OR gates. The start-up was called Neural Semiconductor. I started into a couple of parts and placement designs for them, but they folded before I could get paid or do detailed design.

Fortunately for both of us, Stan found a new job at Oricon in San Diego where he found funding to pursue his idea. The above board is a VME design I did for Stan at Orincon that employed 22 of the chips he designed. It all seemed to work well. But it being a defense oriented company, I never heard any more about it. Stan did not stay there long either.