Decoding the NEC V20 Microcode
Summary
The article documents the process of decoding the NEC V20 microcode to enable cycle-accurate emulation. It describes obtaining die photography of a V20, extracting the microcode ROM bitmap, using MaskRomTool to map bit positions, and training a convolutional neural network (CNN) to classify bits as 0 or 1. It also covers decoding the Group Decode ROM PLA, interpreting the 29-bit microcode words, and how the V20/V30 metal-layer switch affects microcode masking. The piece emphasizes collaborative reverse-engineering work and points to GitHub artifacts for further exploration.