MOSFET 1962 Step 15 Die Yield Economics Bipolar Comparison
# Measure the Yield Economics: Why a Simpler Device Did Not Mean a Cheaper One Yet
## 1. Fewer Process Steps, Not Yet Fewer Defects
A MOS transistor needs fewer photomasking and diffusion steps than a bipolar transistor of comparable function, which should mean higher yield and lower cost -- and in 1962-63 it did not, because the oxide-charge and mobile-ion sensitivities this series' Steps 7 and 14 describe made MOS yield highly dependent on contamination control bipolar lines had already spent a decade perfecting. Early commercial MOS yields ran well below contemporary bipolar yields on comparable die sizes, which meant the simpler device was, for its first several years on the market, often the more expensive one per working unit.
## 2. The Economics That Eventually Flipped
The yield gap closed over the following several years as oxide-cleanliness practices (the same practices named in Step 13's account of Fairchild's process-control program) matured industry-wide, and MOS's fewer-step cost advantage began to show up in actual unit economics by the mid-to-late 1960s -- the same period this project's siliconGate_1967 series covers for a different, specific process improvement. This step closes on the honest 1962-63 picture only: a technically simpler device that was not yet the cheaper one.