1. Purpose of the gate
Concept proven is the first technical gate and the evidence base for the seed round. It decides whether the company's core technical claim, the advantage on which the thesis rests, has been shown with enough rigour that a sceptical outside engineer would accept it. A demo that impresses in a meeting is not enough; the results must be reproducible, measured on representative workloads and written up with their method and limits.
The gate criteria are an architecture model that runs representative workloads and reproduces the claimed advantage, an FPGA prototype that demonstrates the key function on real inputs, a proof of concept report with results, method and limits written up, a statement of how prototype results translate to a chip on the selected node with the main uncertainties, and invention disclosures recorded for anything new.
The silicon translation criterion deserves particular attention. An FPGA or emulator shows function and relative behaviour, but its clock rates, power and area say little directly about a custom chip. The review must see a reasoned argument for what the advantage becomes on silicon, which assumptions it rests on and which of them only a test chip can settle.
10 more sections for registered members.