[CMOS]

Originally Posted by einzlr

That's what I understand, too, but they seem to be saying that the software was assuming much tighter tolerances than actually exist, so when it observes values outside those assumptions it concludes that something has broken when in fact it has not.

The way I would read that is they programmed the ECU to be used on a high efficiency engine with tight tolerances then swapped parts for cheap ones to save a $1 and forgot that the computer was programmed by the person who actually knew what the engine is capable of through testing and that spinning garbage at 7400RPM would raise a few red flags.